- Email: [email protected]
SEDATION FOR ENDOSCOPY
Br. J. Anaesth. (1989), 63, 244-247
CORRESPONDENCE
A. G.JENSEN P. A. HANSEN
Esbjerg, Denmark REFERENCES 1. Lambert MJ. Air embolis in central venous catheterization: Diagnosis, treatment and prevention. Southern Medical Journal 1982; 10: 1189-1191.
2. Michenfelder JD, Miller RH, Gronert GA. Evaluation of an ultrasonic device for diagnosis of venous air embolism. Anesthesiology 1972; 2: 164-167. 3. Gronert GA, Messick J, Cucchiari RF, Michenfelder JD. Paradoxical air embolus from a patent foramen ovale. Anesthesiology 1979; 50: 548-549. 4. Michel L, Poskanzer DC, McKusick KA, Nardi GL, Malt RA. Fatal paradoxical air embolus to the brain. Journal of Parenteral and Enleral Nutrition 1982; 1: 68-70. 5. Perkins-Pearson N, Marshall WK, Dedford RF. Atrial pressures in the seated position. Anesthesiology 1982; 57: 493-^97. 6. Metha M, Sokoll MD. Relation of right and left atrial pressure during venous air embolism. Anesthesiology 1981; 55: A238. 7. Adornato DC, Gildenberg PL, Ferrario CM, Smart J, Frost EAM. Pathophysiology of venous air embolus in dogs. Anesthesiology 1978; 49: 120-127. 8. Diethrich EB, Koopot R, Maze A, Dyess N. Successful reversal of brain damage from iatrogenic air embolism. Surgery, Gynaecology and Obstetrics 1982; 154: 572-575. 9. Khalil SN, Madan V, Rigor BM, Fields WS, Unger KM. Systemic air embolism following induction of artificial pneumothorax under anaesthesia, with successful management. British Journal of Anaesthesia 1979; 51: 461-463.
SEDATION FOR ENDOSCOPY Sir,—We note with concern the paper by Boldy and coworkers on sedation for endoscopy [1]. In the results section, the authors state that "There were no cardiorespiratory problems in either group." There is no mention of any monitoring used in this study or at what times measurements or observations were made. Cardiorespiratory changes during sedation for endoscopy are well known. Several studies have shown ECG changes during sedation and insertion of an endoscope [2-5], and significant reductions in oxygen tension have been demonstrated after introduction of the endoscope alone, after sedation with a benzodiazepine and after a combination of benzodiazepine and opioid [6-11]. Midazolam alone has been shown to cause marked respiratory depression [12]. We have completed a study of cardiorespiratory changes in 20 patients undergoing prolonged endoscopy. Data were recorded continuously by an Atari 1040 ST micro-computer, from ECG lead CM5, a pulse oximeter (Ohmeda Biox 3700), and a non-invasive arterial pressure monitor (Datascope). Sedation was provided by pethidine 25-50 mg and midazolam titrated to effect (mean dose 8 mg). The study commenced before administration of the sedative and continued for the first 1 h of recovery. Oxygen saturation decreased in all patients (min 82 %, SEM 12.5%), remained so for the duration of the examination and persisted into the recovery period. At the end of the study, saturation had not returned to baseline in 11 patients. An
Downloaded from http://bja.oxfordjournals.org/ at Karolinska Institutet on July 22, 2015
NON-FATAL PARADOXICAL AIR EMBOLISM Sir,—Cerebral air emboli are usually fatal or may cause serious neurological conditions. The risk of such an embolus during insertion of a central venous catheter (CVC) is well recognized [1]. We report a patient in whom an embolus occurred when a three-way tap on a CVC became accidentally disconnected. The patient was a 33-yr-old man with acute myeloid leukaemia who was receiving treatment via a permanently located CVC in the right internal jugular vein. The leukaemia was in good remission and the patient was cutting the lawn when he suddenly felt unwell, with sweating, dizziness and palpitations. The patient's wife heard a buzzing sound under his shirt and thought it was a bee. He then collapsed and had convulsions accompanied by blood streaming from the CVC. On reaching the casualty department, it was imposible to establish verbal contact with the patient. He had flexion spasms of all extremities every 5-10 s, more marked on the left. Both pupils were dilated moderately, equal and reactive to light. Bilateral foot clonus was observed and his heart rate was 140 beat min"1. Arterial blood-gas analysis revealed slight hypoxia with PaOl 8.6 kPa, and 5aOj 90%. The patient was treated with oxygen, a left lateral tilt, Diazemuls 45 mg i.v. and phenobarbitone 200 mg i.v. The seizures stopped, but no response could be obtained from the patient. After observation for 5 h he awoke suddenly and was neurologically completely intact. Venous air embolism with CV catheters can occur when a subject is in an upright position. This has been demonstrated in 30-40% of patients who undergo neurosurgery in the sitting position [1]. The surest method of determining the diagnosis is by Doppler ultrasound [2]. Our patient presented with the signs of a cerebrally-triggered attack; air may have entered the cerebral circulation, possibly via a patent foramen ovale, which is present in 20-30% of the population. Previous instances of paradoxical air embolus via the foramen ovale have been described [3,4], and a pressure gradient of only 4 mm Hg is necessary to produce a right-to-left intracardiac shunt [5]. It has been shown in animals that the presence of air alone in the right atrium is enough to cause such a shunt [6], and this is seen especially when a bolus of air is involved [7]. Cerebral air emboli often result in serious neurological conditions [4], but in this patient there were no sequelae. Should brain damage occur, the treatment is limited. One patient has been described [8] in whom treatment with barbiturate-induced coma, slight hypothermia and hyperventilation were thought to have been effective. Hyperbaric oxygen treatment has also been described [9]. In the present patient complete recovery occurred without treatment.
CORRESPONDENCE
245
of 42 % was recorded in one patient. Sixteen of the 20 patients developed tachycardia. Ten patients developed supraventricular ectopic beats, ventricular ectopic beats or both. ECG changes resolved during the recovery period. A significant correlation was found between the occurrence of S-T segment depression and hypoxia (r = 0.818, P < 0.00005). No correlation was found between S-T segment depression and arterial pressure, heart rate or rate-pressure product. From the results of our study and those of others, cardiorespiratory monitoring would appear to be mandatory during upper gastrointestinal endoscopy, especially if opioid analgesics are administered, hypoxia exists already or the patient is in a high risk group. A. W. MURRAY G. KENNY
REFERENCES 1. Boldy DAR, Lever LR, Unwin PR, Spencer PA, Hoare AM. Sedation for endoscopy: midazolam or diazepam and pethidine. British Journal of Anaesthesia 1988; 61: 698-701. 2. De Masi CJ, Akdamar K. Electrocardiography during upper gastrointestinal endoscopy. Gastrointestinal Endoscopy 1969; 16: 33-34. 3. Pyorala K, Salmi HJ, Jussila J, Heikkila J. Electrocardiographical changes during gastroscopy. Endoscopy 1973; 5: 186-193. 4. Fujita R, Kumura F. Arrhythmias and ischaemic changes of the heart induced by gastric endoscopic procedures. American Journal of Gastroenterology 1974; 64: 44-48. 5. Sturges HF, Krone CL. Cardiovascular stress of peroral gastrointestinal endoscopy. Gastrointestinal Endoscopy 1973; 19: 119-122. 6. Zsigmond EK, Flynn K, Martinez OA. Diazepam and meperidine and arterial blood gases in healthy volunteers. Journal of Clinical Pharmacology 1974; 14: 377-381. 7. Zsigmond EK, Shively JG, Flynn K. Diazepam and meperidine and arterial blood gases in patients with chronic obstructive pulmonary disease. Journal of Clinical Pharmacology 1975; IS: 464-^69. 8. Rozen P, Fireman Z, Gilat T. Arterial oxygen tension changes in elderly patients undergoing upper gastrointestinal endoscopy II. Influence of the narcotic R. C. PEARSON premedication and endoscope diameter. Scandinavian R. F. MCCLOY Journal of Gastroenterology 1981; 16: 299-303. Manchester 9. Rozen P, Fireman Z, Gilat T. The causes of hypoxaemia in elderly patients during endoscopy. Gastrointestinal REFERENCES Endoscopy 1982; 28: 243-246. 1. Boldy DAR, Lever LR, Unwin PR, Spencer PA, Hoare 10. Pecora AA, Chiesa JC, Alloy AM, Santoro J, Lazarus B. AM. Sedation for endoscopy: midazolam or diazepam and The effect of upper gastro-intestinal endoscopy on arterial pethidine? British Journal of Anaesthesia 1988; 61: oxygen tension in smokers and non-smokers with and 698-701. without premedication. Gastrointestinal Endoscopy 1984; 2. Whitwam JG, Al-Khudhairi D, McCloy RF. Com30: 284-288. parison of midazolam and diazepam in doses of com11. Whorwell PJ, Smith CL, Foster KJ. Arterial blood gas parable potency during gastroscopy. British Journal of tensions during upper gastro-intestinal endoscopy. Gut Anaesthesia 1983; 55: 773-777. 1976; 17: 797-800. 3. Bell GD, Reeve PA, Moshiri M, Morden A, Coady T, 12. Bell GD, Reeve PA, Moshiri M, Morden A, Coady T, Stapleton PJ, Logan RFA. Intravenous midazolam: a Stapleton PJ, Logan RFA. Intravenous midazolam: A study of the degree of oxygen saturation occurring during study of the degree of oxygen desaturation occurring upper gastrointestinal endoscopy. British Journal of during upper gastrointestinal endoscopy. British Journal Clinical Pharmacology 1987; 23: 703-708. of Clinical Pharmacology 1987; 23: 703-708.
Downloaded from http://bja.oxfordjournals.org/ at Karolinska Institutet on July 22, 2015
Glasgow
Sir,—We read with concern the paper by Boldy and colleagues [1], which seems to highlight much of the misunderstanding concerning the use of i.v. sedation. The authors do not provide details of patient weight, making it difficult to be precise concerning doses used. If we assume that the mean male weight was approximately 70 kg, the mean dose of midazolam used in males was 0.16 mg kg""1. The manufacturer's recommended dose is 0.07 mg kg"1 for sedation during endoscopy—a dose found to be satisfactory by the second trial from the Hammersmith group [2] (not quoted by Boldy). As midazolam has approximately twice the hypnotic activity of diazepam, it seems unreasonable to compare diazepam 0.15 mg kg"1 with midazolam 0.16 mg kg"1, and hardly surprising that, after endoscopy, patients were more sedated after midazolam. Five milligrams is the maximum dose of midazolam we use, as the average patient is too profoundly sedated to be cooperative after being given midazolam 10 mg i.v. Bell and colleagues [3,4] showed significant decreases in oxygen saturation (to less than 80%) during sedation with a benzodiazepine, exacerbated by passage of the gastroscope, and although they used larger doses (means of 6.7 and 6.3 mg) than we would recommend, these were considerably short of the doses of midazolam used by Boldy and colleagues. The addition of opioids to benzodiazepines compounds respiratory depression and hypoxia [5]. We believe that opioid supplementation has no place in sedation for routine upper gastrointestinal endoscopy. As the peak effect of benzodiazepines given as a single i.v. injection occurs after 2-3 min, "titration" by injecting to a given endpoint cannot take this into account unless this time period is observed between each incremental dose. The pharmacokinetics of midazolam make it better suited than diazepam to sedation for endoscopy and its superior amnesic properties provide an added clinical benefit. The recent availability of flumazenil, a specific benzodiazepine antagonist, obviates Boldy's assertions that midazolam-induced amnesia is a disadvantage. Antagonism of midazolam 5 mg with flumazenil 0.5 mg i.v. prevents further amnesia, reverses sedation and returns psychomotor variables to normal within 1 min [6]. In all published trials, there has been no clinically significant residual sedation caused by the disparity between the half-lives of the two drugs (57 min cf. 1.3-2.2 h). Of course, when excessive doses of midazolam are used in individual patients, this possibility may be increased.
CORRESPONDENCE
A. G.JENSEN P. A. HANSEN
Esbjerg, Denmark REFERENCES 1. Lambert MJ. Air embolis in central venous catheterization: Diagnosis, treatment and prevention. Southern Medical Journal 1982; 10: 1189-1191.
2. Michenfelder JD, Miller RH, Gronert GA. Evaluation of an ultrasonic device for diagnosis of venous air embolism. Anesthesiology 1972; 2: 164-167. 3. Gronert GA, Messick J, Cucchiari RF, Michenfelder JD. Paradoxical air embolus from a patent foramen ovale. Anesthesiology 1979; 50: 548-549. 4. Michel L, Poskanzer DC, McKusick KA, Nardi GL, Malt RA. Fatal paradoxical air embolus to the brain. Journal of Parenteral and Enleral Nutrition 1982; 1: 68-70. 5. Perkins-Pearson N, Marshall WK, Dedford RF. Atrial pressures in the seated position. Anesthesiology 1982; 57: 493-^97. 6. Metha M, Sokoll MD. Relation of right and left atrial pressure during venous air embolism. Anesthesiology 1981; 55: A238. 7. Adornato DC, Gildenberg PL, Ferrario CM, Smart J, Frost EAM. Pathophysiology of venous air embolus in dogs. Anesthesiology 1978; 49: 120-127. 8. Diethrich EB, Koopot R, Maze A, Dyess N. Successful reversal of brain damage from iatrogenic air embolism. Surgery, Gynaecology and Obstetrics 1982; 154: 572-575. 9. Khalil SN, Madan V, Rigor BM, Fields WS, Unger KM. Systemic air embolism following induction of artificial pneumothorax under anaesthesia, with successful management. British Journal of Anaesthesia 1979; 51: 461-463.
SEDATION FOR ENDOSCOPY Sir,—We note with concern the paper by Boldy and coworkers on sedation for endoscopy [1]. In the results section, the authors state that "There were no cardiorespiratory problems in either group." There is no mention of any monitoring used in this study or at what times measurements or observations were made. Cardiorespiratory changes during sedation for endoscopy are well known. Several studies have shown ECG changes during sedation and insertion of an endoscope [2-5], and significant reductions in oxygen tension have been demonstrated after introduction of the endoscope alone, after sedation with a benzodiazepine and after a combination of benzodiazepine and opioid [6-11]. Midazolam alone has been shown to cause marked respiratory depression [12]. We have completed a study of cardiorespiratory changes in 20 patients undergoing prolonged endoscopy. Data were recorded continuously by an Atari 1040 ST micro-computer, from ECG lead CM5, a pulse oximeter (Ohmeda Biox 3700), and a non-invasive arterial pressure monitor (Datascope). Sedation was provided by pethidine 25-50 mg and midazolam titrated to effect (mean dose 8 mg). The study commenced before administration of the sedative and continued for the first 1 h of recovery. Oxygen saturation decreased in all patients (min 82 %, SEM 12.5%), remained so for the duration of the examination and persisted into the recovery period. At the end of the study, saturation had not returned to baseline in 11 patients. An
Downloaded from http://bja.oxfordjournals.org/ at Karolinska Institutet on July 22, 2015
NON-FATAL PARADOXICAL AIR EMBOLISM Sir,—Cerebral air emboli are usually fatal or may cause serious neurological conditions. The risk of such an embolus during insertion of a central venous catheter (CVC) is well recognized [1]. We report a patient in whom an embolus occurred when a three-way tap on a CVC became accidentally disconnected. The patient was a 33-yr-old man with acute myeloid leukaemia who was receiving treatment via a permanently located CVC in the right internal jugular vein. The leukaemia was in good remission and the patient was cutting the lawn when he suddenly felt unwell, with sweating, dizziness and palpitations. The patient's wife heard a buzzing sound under his shirt and thought it was a bee. He then collapsed and had convulsions accompanied by blood streaming from the CVC. On reaching the casualty department, it was imposible to establish verbal contact with the patient. He had flexion spasms of all extremities every 5-10 s, more marked on the left. Both pupils were dilated moderately, equal and reactive to light. Bilateral foot clonus was observed and his heart rate was 140 beat min"1. Arterial blood-gas analysis revealed slight hypoxia with PaOl 8.6 kPa, and 5aOj 90%. The patient was treated with oxygen, a left lateral tilt, Diazemuls 45 mg i.v. and phenobarbitone 200 mg i.v. The seizures stopped, but no response could be obtained from the patient. After observation for 5 h he awoke suddenly and was neurologically completely intact. Venous air embolism with CV catheters can occur when a subject is in an upright position. This has been demonstrated in 30-40% of patients who undergo neurosurgery in the sitting position [1]. The surest method of determining the diagnosis is by Doppler ultrasound [2]. Our patient presented with the signs of a cerebrally-triggered attack; air may have entered the cerebral circulation, possibly via a patent foramen ovale, which is present in 20-30% of the population. Previous instances of paradoxical air embolus via the foramen ovale have been described [3,4], and a pressure gradient of only 4 mm Hg is necessary to produce a right-to-left intracardiac shunt [5]. It has been shown in animals that the presence of air alone in the right atrium is enough to cause such a shunt [6], and this is seen especially when a bolus of air is involved [7]. Cerebral air emboli often result in serious neurological conditions [4], but in this patient there were no sequelae. Should brain damage occur, the treatment is limited. One patient has been described [8] in whom treatment with barbiturate-induced coma, slight hypothermia and hyperventilation were thought to have been effective. Hyperbaric oxygen treatment has also been described [9]. In the present patient complete recovery occurred without treatment.
CORRESPONDENCE
245
of 42 % was recorded in one patient. Sixteen of the 20 patients developed tachycardia. Ten patients developed supraventricular ectopic beats, ventricular ectopic beats or both. ECG changes resolved during the recovery period. A significant correlation was found between the occurrence of S-T segment depression and hypoxia (r = 0.818, P < 0.00005). No correlation was found between S-T segment depression and arterial pressure, heart rate or rate-pressure product. From the results of our study and those of others, cardiorespiratory monitoring would appear to be mandatory during upper gastrointestinal endoscopy, especially if opioid analgesics are administered, hypoxia exists already or the patient is in a high risk group. A. W. MURRAY G. KENNY
REFERENCES 1. Boldy DAR, Lever LR, Unwin PR, Spencer PA, Hoare AM. Sedation for endoscopy: midazolam or diazepam and pethidine. British Journal of Anaesthesia 1988; 61: 698-701. 2. De Masi CJ, Akdamar K. Electrocardiography during upper gastrointestinal endoscopy. Gastrointestinal Endoscopy 1969; 16: 33-34. 3. Pyorala K, Salmi HJ, Jussila J, Heikkila J. Electrocardiographical changes during gastroscopy. Endoscopy 1973; 5: 186-193. 4. Fujita R, Kumura F. Arrhythmias and ischaemic changes of the heart induced by gastric endoscopic procedures. American Journal of Gastroenterology 1974; 64: 44-48. 5. Sturges HF, Krone CL. Cardiovascular stress of peroral gastrointestinal endoscopy. Gastrointestinal Endoscopy 1973; 19: 119-122. 6. Zsigmond EK, Flynn K, Martinez OA. Diazepam and meperidine and arterial blood gases in healthy volunteers. Journal of Clinical Pharmacology 1974; 14: 377-381. 7. Zsigmond EK, Shively JG, Flynn K. Diazepam and meperidine and arterial blood gases in patients with chronic obstructive pulmonary disease. Journal of Clinical Pharmacology 1975; IS: 464-^69. 8. Rozen P, Fireman Z, Gilat T. Arterial oxygen tension changes in elderly patients undergoing upper gastrointestinal endoscopy II. Influence of the narcotic R. C. PEARSON premedication and endoscope diameter. Scandinavian R. F. MCCLOY Journal of Gastroenterology 1981; 16: 299-303. Manchester 9. Rozen P, Fireman Z, Gilat T. The causes of hypoxaemia in elderly patients during endoscopy. Gastrointestinal REFERENCES Endoscopy 1982; 28: 243-246. 1. Boldy DAR, Lever LR, Unwin PR, Spencer PA, Hoare 10. Pecora AA, Chiesa JC, Alloy AM, Santoro J, Lazarus B. AM. Sedation for endoscopy: midazolam or diazepam and The effect of upper gastro-intestinal endoscopy on arterial pethidine? British Journal of Anaesthesia 1988; 61: oxygen tension in smokers and non-smokers with and 698-701. without premedication. Gastrointestinal Endoscopy 1984; 2. Whitwam JG, Al-Khudhairi D, McCloy RF. Com30: 284-288. parison of midazolam and diazepam in doses of com11. Whorwell PJ, Smith CL, Foster KJ. Arterial blood gas parable potency during gastroscopy. British Journal of tensions during upper gastro-intestinal endoscopy. Gut Anaesthesia 1983; 55: 773-777. 1976; 17: 797-800. 3. Bell GD, Reeve PA, Moshiri M, Morden A, Coady T, 12. Bell GD, Reeve PA, Moshiri M, Morden A, Coady T, Stapleton PJ, Logan RFA. Intravenous midazolam: a Stapleton PJ, Logan RFA. Intravenous midazolam: A study of the degree of oxygen saturation occurring during study of the degree of oxygen desaturation occurring upper gastrointestinal endoscopy. British Journal of during upper gastrointestinal endoscopy. British Journal Clinical Pharmacology 1987; 23: 703-708. of Clinical Pharmacology 1987; 23: 703-708.
Downloaded from http://bja.oxfordjournals.org/ at Karolinska Institutet on July 22, 2015
Glasgow
Sir,—We read with concern the paper by Boldy and colleagues [1], which seems to highlight much of the misunderstanding concerning the use of i.v. sedation. The authors do not provide details of patient weight, making it difficult to be precise concerning doses used. If we assume that the mean male weight was approximately 70 kg, the mean dose of midazolam used in males was 0.16 mg kg""1. The manufacturer's recommended dose is 0.07 mg kg"1 for sedation during endoscopy—a dose found to be satisfactory by the second trial from the Hammersmith group [2] (not quoted by Boldy). As midazolam has approximately twice the hypnotic activity of diazepam, it seems unreasonable to compare diazepam 0.15 mg kg"1 with midazolam 0.16 mg kg"1, and hardly surprising that, after endoscopy, patients were more sedated after midazolam. Five milligrams is the maximum dose of midazolam we use, as the average patient is too profoundly sedated to be cooperative after being given midazolam 10 mg i.v. Bell and colleagues [3,4] showed significant decreases in oxygen saturation (to less than 80%) during sedation with a benzodiazepine, exacerbated by passage of the gastroscope, and although they used larger doses (means of 6.7 and 6.3 mg) than we would recommend, these were considerably short of the doses of midazolam used by Boldy and colleagues. The addition of opioids to benzodiazepines compounds respiratory depression and hypoxia [5]. We believe that opioid supplementation has no place in sedation for routine upper gastrointestinal endoscopy. As the peak effect of benzodiazepines given as a single i.v. injection occurs after 2-3 min, "titration" by injecting to a given endpoint cannot take this into account unless this time period is observed between each incremental dose. The pharmacokinetics of midazolam make it better suited than diazepam to sedation for endoscopy and its superior amnesic properties provide an added clinical benefit. The recent availability of flumazenil, a specific benzodiazepine antagonist, obviates Boldy's assertions that midazolam-induced amnesia is a disadvantage. Antagonism of midazolam 5 mg with flumazenil 0.5 mg i.v. prevents further amnesia, reverses sedation and returns psychomotor variables to normal within 1 min [6]. In all published trials, there has been no clinically significant residual sedation caused by the disparity between the half-lives of the two drugs (57 min cf. 1.3-2.2 h). Of course, when excessive doses of midazolam are used in individual patients, this possibility may be increased.