Capnographic Monitoring of Respiratory Activity Improves Safety of Sedation for Endoscopic Cholangiopancreatography and Ultrasonography

Article Outline

• Abstract
• Patients and Methods
  • Study Population
  • Study Design
  • Intervention
  • Randomization
  • Patient Demographics and Procedure Variables
  • Study Procedure
  • Study Outcome
  • Sample Size Estimations
  • Statistical Analysis
• Results
  • Patients
  • Outcomes
  • Relationship Between Hypoxemia and Ventilation
  • Relationship Between Hypoxemia and Other Clinical Variables
  • Effect of Obesity on Hypoxemia and Apnea
  • Serious Adverse Events
• Discussion
• Acknowledgments
• References
• Copyright

Background & Aims

The Joint Commission on the Accreditation of Healthcare Organizations recommends ventilation monitoring during procedural sedation for gastrointestinal endoscopy. We sought to determine whether intervention, based on a microstream capnography-based ventilation monitoring system that has been shown to function as an early warning system for hypoxemia, would decrease hypoxemia during endoscopy.

Methods

Subjects undergoing elective endoscopic retrograde cholangiopancreatography (ERCP) and endoscopic ultrasonography (EUS) under procedural sedation with a combination of opioid and benzodiazepine were randomly assigned to either a study arm in which the endoscopy team was blinded to capnography or an open arm in which the endoscopy team was prompted of capnographic changes. The primary end point was the occurrence of hypoxemia; secondary end points were the occurrences of severe hypoxemia, apnea, and oxygen supplementation.

Results

A total of 263 subjects were enrolled; 247 were analyzed for efficacy. The numbers of hypoxemic events in the blinded and open arms were 132 and 69, respectively (P < .001). Thirty-five percent of all hypoxemic events occurred with completely normal ventilation. Hypoxemia developed in 69% of patients in the blinded arm compared with 46% in the open arm (P < .001). Severe hypoxemia percentages in the blinded and open arms were 31% and 15% (P = .004), for apnea were 63% and 41% (P < .001), for oxygen supplementation were 67% and 52% (P = .02), and for recurrent hypoxemia after oxygen supplementation were 38% and 18% (P = .01), respectively.

Conclusions

Capnographic monitoring of respiratory activity improves patient safety during procedural sedation for elective ERCP/EUS by reducing the frequency of hypoxemia, severe hypoxemia, and apnea.

Abbreviations used in this paper: ASA, American Society of Anesthesiologists, CI, confidence interval, ERCP, endoscopic retrograde cholangiopancreatography, EUS, endoscopic ultrasonography, HR, hazard ratio

Procedural sedation administered for gastrointestinal endoscopy has 2 cardinal goals: patient safety and comfort.¹, ² Better understanding of the pathophysiology of sedation and patient monitoring has contributed significantly to improved safety and comfort.¹, ², ³ Similar achievements in anesthesia were highlighted in the Institute of Medicine report To Err is Human, in which the use of monitoring devices and better understanding of pathophysiology were credited with significantly reducing mortality.⁴ The Joint Commission on the Accreditation of Healthcare Organizations recommends “respiratory frequency and adequacy of pulmonary ventilation should be continually monitored in patients undergoing conscious sedation.”⁵ Despite these recommendations, dedicated ventilation monitoring during procedural sedation is not routinely performed either due to the expense of hiring trained observers dedicated exclusively to monitoring ventilation or lack of enthusiasm for automated ventilation monitoring devices, which have been shown to be superior to clinical monitoring.⁶

All automated ventilation monitors utilize the principles of capnography first discovered in 1943 by Karl Luft, who found that the carbon dioxide molecules absorb specific wavelengths of infrared light, with their absorption patterns reflecting their relative concentration in the given milieu.⁷ Previous capnography systems using transcutaneous technique did not attain popularity due to variable performance, training, and cost.⁸ Capnography systems utilizing microstream technology use the physiologic differences in the carbon dioxide concentration in the respired air to display respiratory activity as crests (expiration) and troughs (inspiration).⁹ Except for one case series that suggested that microstream capnography monitoring is more sensitive for detecting respiratory abnormalities compared with trained nurses and could function as an early warning system for hypoxemia,⁶ and a pediatric randomized study that found fewer hypoxemic events with electronic monitoring,¹⁰ there are, to our knowledge, no outcome studies with this ventilation monitoring device in adult gastrointestinal endoscopy.

Therefore, we assessed the efficacy of interventions based on capnographic monitoring of respiratory activity and conducted a randomized controlled trial to reduce hypoxemia in adult patients undergoing elective endoscopic retrograde cholangiopancreatography (ERCP) or endoscopic ultrasonography (EUS). The primary study outcome was the proportion of patients developing hypoxemia, defined as oxygen saturation of <90% for ≥15 seconds, in the 2 groups.

Back to Article Outline

Patients and Methods 

Study Population 

The study protocol was reviewed and approved by the institutional review board of the Cleveland Clinic (Cleveland, OH). Patient enrollment began in January 2007 and concluded in May 2008. All consecutive patients during the investigators' (M.A.Q., P.A.T.) research days presenting for elective inpatient and outpatient ERCP or EUS to our endoscopy unit were considered for enrollment if they had all of the following inclusion criteria: (1) age 18 years or older, (2) American Society of Anesthesiologists (ASA) class 1–3, and (3) ability to give written informed consent. Patients were excluded from enrollment if they met any of the following criteria: (1) ASA class 4 or 5, (2) unable to give informed consent, (3) required emergency procedures, (4) required monitored anesthesia care sedation, (5) used oxygen or noninvasive ventilation devices such as continuous positive airway pressure or bilevel positive airway pressure before the procedure, or (6) had allergies to fentanyl, meperidine, or midazolam.

Study Design 

The study was a single-center, randomized, double-blind, parallel-assignment trial. We assessed the efficacy of capnographic monitoring of respiratory activity for reducing the proportion of patients developing hypoxemia by randomly assigning subjects to an open arm or a blinded arm. One of the 2 independent observers (M.A.Q., P.A.T.) monitored the capnographic display and prompted the endoscopy team of the changes in respiratory activity as follows. In the open arm, the independent observer signaled all respiratory abnormalities detected on the capnography monitor within 5–10 seconds of onset. Abnormalities included flat line for >5 seconds or >75% reduction in amplitude of respiratory waves compared with baseline for more than 5 seconds and/or respiratory rate of <8/min. In the blinded arm, no signals were given to the endoscopy team for abnormal ventilation regardless of duration, but the only signal consisted of apnea lasting more than 30 seconds as mandated by our institutional review board out of safety concern for continued blinding in the presence of prolonged apnea. Thus, interventions were initiated in both groups based on capnography, albeit at different intervals (5–10 seconds in the open arm for abnormal ventilation and >30 seconds in the blinded arm for apnea). The institutional review board did realize that such an intervention would potentially decrease hypoxemia in the blinded arm by capnography-based interventions and bias the study in favor of the blinded arm, but, understandably, considered patient safety to be more important. The endoscopy team (physicians and nurses) was unaware of the capnographic findings at any given time in both arms because the only signal consisted of a verbal cue “not breathing properly.”

Intervention 

The interventions consisted of (1) patient stimulation, (2) withholding medications, and/or (3) oxygen supplementation. These interventions were triggered by 2 circumstances: (1) being alerted by the independent observer regardless of oxygen saturations or (2) declining oxygen saturations noticed by the endoscopy team regardless of any alerts. All interventions were decided entirely by the endoscopy team, which was blinded to the randomization arm. Oxygen supplementation was not used either before or after the procedure if the saturations remained >90%.

Randomization 

The randomization process was generated by the Department of Biostatistics at our institution and consisted of variable blocks ranging from 4 to 8 in concealed envelopes. The participants were recruited by the study nurse (P.A.T.) or independent observer (M.A.Q.). The patient and the endoscopy team were blinded to the allocation assignment. The independent observers did not directly take part in the randomization allocation process but knew about the randomization assignment before the start of the procedure.

Patient Demographics and Procedure Variables 

We obtained a number of patient variables, including age; sex; body mass index; history of smoking; concurrent use of benzodiazepines or narcotic medications; history of diseases of the heart, lung, kidney, or liver; sleep apnea; and ASA status. The procedural variables obtained were whether the patients were outpatients or inpatients, if trainees participated during the procedure, baseline oxygen saturation, type of procedure (ERCP or EUS), endoscopist performing the procedure, doses and timing of medications during the procedure, and duration of the procedure.

Study Procedure 

All patients underwent standard monitoring with continuous display of heart rate and pulse oximetry along with blood pressure every 5 minutes. We used specialized bite blocks (Smart BitebloCO₂; Oridion Capnography, Inc, Needham, MA) consisting of a mouthpiece for the passage of an endoscope and a long plastic tube with 2 nasal prongs and 1 oral prong for capturing nasal and oral breathing, respectively. The plastic tube was connected to the capnographic device (Capnostream 20; Oridion Capnography, Inc), which displays respiratory activity as crests (expiration) and troughs (inspiration). All patients had run-in of several minutes before the first dose of medications was given, which signaled the beginning of the procedure. All abnormal events detected on pulse rate, pulse oximetry, and capnography were crosschecked for any mechanical issues related to devices and sensors. All patients were sedated with midazolam in combination with meperidine or fentanyl. Diazepam was added when patients were difficult to sedate with the previously described combination. The staff physician determined the doses of medications based on a previously reported protocol.¹¹ The initial doses of medications ranged from 25 to 75 U of meperidine or fentanyl along with 2 mg of midazolam based on the patient's age, body mass index, and concurrent use of narcotics/benzodiazepines. Increments of 25 U of meperidine (milligrams) or fentanyl (micrograms) with or without 1 mg of midazolam were administered when patient discomfort or lighter levels of sedation was noticed. All vital signs, including blood pressure, oxygen saturation, and pulse rate, were checked before administering any sedation medications. Capnography was not checked, but the endoscopy team was informed of abnormalities according to the protocol described previously. The final extubation was considered the end of the procedure. The patients were monitored for 3–5 minutes and then transferred to the recovery area. They were discharged from the endoscopy unit once they attained full recovery.

Study Outcome 

The principal study outcome was hypoxemia, defined as oxygen saturation of <90% for ≥15 seconds. The secondary outcomes included the following: (1) severe hypoxemia (oxygen saturation ≤85% regardless of duration; however, as with other outcome assessments, all sensors were checked before labeling any episode as severe hypoxemia to avoid mislabeling any episode due to mechanical problems), (2) requirement of supplemental oxygen (concurrent oxygen use was an exclusion criteria), (3) apnea (absence of respiratory activity denoted by a flat line on the capnometer for ≥15 seconds), and (4) abnormal ventilation (flat line for ≥5 seconds but <15 seconds, >75% reduction in amplitude of respiratory waves for ≥5 seconds). The ventilation was considered normal if there was no change from baseline pattern or minimal change that did not fit into any of the above categories in either the amplitude or frequency of respiratory cycles. Two persons assessed hypoxemia and severe hypoxemia independently and resolved any differences by consensus: the circulating endoscopy nurse who was blinded to the randomization arm and an independent observer who was not blinded to the randomization arm. Apnea and abnormal ventilation were assessed only by the independent observer.

Sample Size Estimations 

The sample size calculations were based on χ² test for comparison of 2 proportions and estimation that a reduction of hypoxemia from 40% of patients to 20% will be clinically relevant. We used a conservative estimate of 40% in the blinded arm even though the proportion of patients developing hypoxemia is reported to be in the range of 44% and 70% in the literature.⁷, ¹¹, ¹², ¹³, ¹⁴, ¹⁵ Based on a reduction from 40% to 20%, we estimated that 250 patients (125 patients in each arm) would be required for an α of 0.05 and power of 90%. Because we performed an unplanned interim analysis at 50% enrollment to present the data at Digestive Disease Week 2008,¹⁶ the sample size was adjusted to 263 patients.

The manufacturer (Oridion Capnography, Inc) provided the capnographic monitor (Capnostream 20) and specialized bite blocks (Smart BitebloCO₂) but had no role in study design, data collection, data analysis, or manuscript preparation.

Statistical Analysis 

Descriptive statistics were computed for all variables to provide means and SDs for continuous variables and frequencies for categorical variables. Intergroup differences were assessed by Student t tests for continuous factors and Pearson's χ² tests for categorical factors, while Wilcoxon rank sum tests were used to assess differences in number of events between the groups. We planned efficacy analysis for hypoxemia, severe hypoxemia, apnea, oxygen use, abnormal ventilation, and recurrent hypoxemia after oxygen supplementation. We also planned a safety analysis of serious adverse events such as death, respiratory failure, or requirement of reversal agents. We performed analyses both by intention-to-treat and per-protocol methods. A time-to-event analysis was performed to assess risk factors for hypoxemia. Follow-up time was defined as time to first hypoxemia event or duration of procedure if subjects did not have hypoxemia episodes during the procedure. Kaplan–Meier plots were constructed and univariable and multivariable Cox proportional hazards models were built to evaluate the factors associated with hypoxemia. Interactions between predetermined clinical variables such as capnography, obesity, diazepam use, and ASA class were assessed and considered for inclusion in the final models. A model containing the capnography group, type of procedure, age, body mass index, meperidine/fentanyl dose, and midazolam dose was created, and a stepwise selection method with 0.35 and 0.1 as the entry and exit significance criteria, respectively, was used to select other variables to be included in the final model. A 2-tailed P value <.05 was considered statistically significant. SAS version 9.1 software (SAS Institute, Cary, NC) and R version 2.4.1 software (The R Institute for Statistical Computing, Vienna, Austria) were used to perform all analyses.

Back to Article Outline

Results 

Patients 

A total of 311 subjects presenting for elective ERCP or EUS were screened for enrollment; 23 refused participation, and another 25 failed to meet eligibility criteria. Of the 263 subjects randomized in the study, 16 were excluded from the efficacy analysis: 11 due to inability to perform procedures, 3 due to protocol deviations because they were already on oxygen before the procedure, and 2 due to duplicate enrollment (only first procedures were considered for analysis). Thus, 247 subjects were included in the efficacy analysis but all 263 in the safety analysis (Figure 1). Table 1 depicts equitable distribution between the capnography blinded and capnography open groups with respect to the baseline demographics of patients and procedural characteristics, with the exception of trainee involvement in the procedure (48% in the blinded arm and 36% in the open arm; P = .046).

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1. 

  Patient flow through the trial.

Table 1. Demographic, Clinical, and Procedural Characteristics of Subjects

Factor	Capnography blinded arm (n = 123)	Capnography open arm (n = 124)	P value
Age (y), mean (SD)	60.6(14.3)	60.8(14.4)	.91
Male sex	62(50.4)	61(49.2)	.85
Body mass index (kg/m2), mean (SD)	26.2(5.6)	26.5(5.8)	.71
Obese (body mass index ≥30 kg/m2)	28(22.8)	26(21.0)	.73
Smoking			.6
Never	80(67.2)	89(72.4)
Current smoker	23(19.3)	22(17.9)
Ex-smoker	16(13.5)	12(9.8)
Regular narcotic/sedative use	30(24.4)	36(29.0)	.41
Regular benzodiazepine use	25(20.3)	21(16.9)	.49
Heart disease	27(22.0)	29(23.4)	.79
Lung disease	13(10.6)	8(6.5)	.25
Renal disease	16(13.0)	14(11.3)	.68
Liver disease	27(22.0)	29(23.4)	.79
Sleep apnea	16(13.0)	14(11.3)	.68
ASA class			.99
1	9(7.3)	9(7.3)
2	86(69.9)	86(69.4)
3	28(22.8)	29(23.4)
Outpatient or ambulatory patients	111(91.0)	111(90.2)	.84
Type of procedure			.92
ERCP	32(26.0)	33(26.6)
EUS	91(74.0)	91(73.4)
Baseline oxygen saturation	97.4(1.9)	97.3(2.1)	.6
Procedural time (min)	34.4(12.5)	37.2(16.1)	.13
Total dose of meperidine (mg) or fentanyl (μg)	127.4(40.7)	126.2(44.4)	.82
Total dose of midazolam (mg)	6.4(2.4)	6.2(2.5)	.54
Total dose of diazepam (mg)	3.3(8.0)	3.5(7.8)	.83
Participation of trainee in procedure	59(48.4)	44(35.8)	.046
Physician			.31
1	22(17.9)	25(20.2)
2	45(36.6)	40(32.3)
3	3(2.4)	11(8.9)
4	7(5.7)	10(8.1)
5	21(17.1)	17(13.7)
6	9(7.3)	5(4.0)
7	16(13.0)	16(12.9)

NOTE. Values are presented as n (%) unless otherwise noted. P values correspond to Student t tests and Pearson's χ² tests as appropriate.

Outcomes 

A total of 85 subjects (69%) from the blinded arm and 57 (46%) from the open arm developed at least one episode of hypoxemia (P < .001) (Table 2). Similar trends were noted for severe hypoxemia (31% and 15%, respectively; P = .004), for apnea (63% and 41%, respectively; P < .001), and for oxygen supplementation (67% and 52%, respectively; P = .02). However, there was no significant difference between the 2 groups for abnormal ventilation (82% and 77%, respectively; P = .29). After oxygen supplementation was started due to hypoxemia, significantly more patients in the blinded arm had recurrent hypoxemia compared with the open arm (38% vs 18%, respectively; P = .01). The statistical significance was maintained for the previously described outcomes on intention-to-treat analyses whether the outcomes in the excluded patients were assumed all positive or all negative for the excluded patients.

Table 2. Proportion of Subjects With Preselected Outcomes in the 2 Groups

Factor	Capnography blinded arm (n = 123)	Capnography open arm (n = 124)	P value
Hypoxemia	85(69.1)	57(46.0)	<.001
Severe hypoxemia	38(30.9)	19(15.3)	.004
Oxygen supplementation	82(66.7)	65(52.4)	.02
Recurrent hypoxemia after oxygen supplementation	31/82(37.8)	10/57(17.5)	.01
Abnormal ventilation	101(82.1)	95(76.6)	.29
Apnea	77(62.6)	51(41.1)	<.001
Hypotension	6(4.9)	10(8.1)	.31
No. of hypoxemia events			<.001
0	38(30.9)	67(54.0)
1	54(43.9)	47(37.9)
2	20(16.3)	8(6.5)
3+	11(8.9)	2(1.6)
No. of hypoxemia events with abnormal ventilation			.015
0	21(24.7)	24(42.1)
1	48(56.5)	28(49.1)
2	9(10.6)	4(7.0)
3+	7(8.2)	1(1.8)
No. of hypoxemia events with normal ventilation			.81
0	49(57.7)	30(52.6)
1	28(32.9)	25(43.9)
2	7(8.2)	1(1.8)
3	1(1.2)	1(1.8)
No. of apneic events			<.001
0	46(37.4)	73(58.9)
1	38(30.9)	33(26.6)
2	25(20.3)	13(10.5)
3+	14(11.4)	5(4.0)

NOTE. Values are presented as n (%). P values correspond to Pearson's χ² tests for occurrence of hypoxemia, major hypoxemia, hypoventilation, apnea, hypotension, and use of o₂ and to Wilcoxon rank sum tests for all number of events factors.

Relationship Between Hypoxemia and Ventilation 

There were a total of 201 hypoxemic events during the study: 132 in the blinded arm and 69 in the open arm, respectively (P < .001) (Table 2). Thirty-five percent of all hypoxemic events occurred with normal ventilation or no change in respiratory activity from baseline (both by capnographic monitoring and nurses' assessments), while the remainder occurred with abnormal ventilation or apnea. Normal ventilation-associated hypoxemia occurred more commonly in the open arm compared with the blinded arm.

Relationship Between Hypoxemia and Other Clinical Variables 

Multivariable Cox proportional hazards model revealed that significant factors for hypoxemia were age (5-year increments; hazard ratio [HR], 1.2; 95% confidence interval [CI], 1.07–1.2), female sex (HR, 1.9; 95% CI, 1.3–2.7), blinded arm of capnography (or current standard of care) (HR, 1.9; 95% CI, 1.4–2.7), EUS procedure (HR, 1.8; 95% CI, 1.2–2.8), and baseline oxygen concentration (1 unit increase) (HR, 0.86; 95% CI, 0.79–0.94) (Table 3). Subgroup analysis revealed that capnographic monitoring was more beneficial during ERCP compared with EUS and in obese patients compared with nonobese patients (Figure 2A–C).

Table 3. Relationship Between Hypoxemia and Preselected Clinical Variables

Factor	Hypoxemia (n = 142)	No hypoxemia (n = 105)	Unadjusted HR (95% CI)a	Adjusted HR (95% CI)a
Age (y), mean (SD)	63.9(12.9)	56.4(15.0)	1.1(1.07–1.2)	1.2(1.07–1.2)
Body mass index (kg/m2), mean (SD)	26.4(5.8)	26.4(5.6)	1.01(0.98–1.03)	1.03(0.99–1.06)
Female sex	77(54.2)	47(44.8)	1.4(0.99–1.9)	1.9(1.3–2.7)
Smoking
Never	100(72.5)	69(66.4)	Reference
Current smoker	21(15.2)	24(23.1)	0.67(0.42–1.07)
Ex-smoker	17(12.3)	11(10.6)	1.2(0.69–1.9)
Outpatient	134(94.4)	88(85.4)	1.9(0.94–3.9)
Regular narcotic/sedative use	33(23.2)	33(31.4)	0.76(0.52–1.1)
Regular benzodiazepine use	28(19.7)	18(17.1)	0.97(0.64–1.5)
Heart disease	39(27.5)	17(16.2)	1.4(0.98–2.0)
Lung disease	15(10.6)	6(5.7)	1.4(0.81–2.4)
Renal disease	16(11.3)	14(13.3)	0.92(0.55–1.6)
Liver disease	24(16.9)	32(30.5)	0.59(0.38–0.92)
Sleep apnea	17(12.0)	13(12.4)	0.92(0.55–1.5)
ASA class
1	8(5.6)	10(9.5)	Reference
2	96(67.6)	76(72.4)	1.5(0.75–3.2)
3	38(26.8)	19(18.1)	2.2(1.03–4.7)
EUS	113(79.6)	69(65.7)	1.6(1.08–2.4)	1.8(1.2–2.8)
Participation of trainee	62(43.7)	41(39.8)	1.1(0.81–1.6)
Baseline oxygen saturation, mean (SD)	97.0(1.9)	97.8(2.0)	0.87(0.80–0.84)	0.86(0.79–0.94)
Procedural time (min), mean (SD)	35.4(14.6)	36.4(14.5)	0.97(0.91–1.03)
Total dose of meperidine (mg) or fentanyl (μg), mean (SD)	126.6(39.1)	127.1(46.9)	0.98(0.89–1.07)	1.2(1.00–1.4)
Total dose of midazolam (mg), mean (SD)	6.0(2.3)	6.6(2.6)	0.92(0.86–0.98)	0.91(0.81–1.02)
Total dose of diazepam (mg), mean (SD)	2.5(6.5)	4.7(9.4)	0.97(0.95–1.00)
Blinded capnography	85(59.9)	38(36.2)	1.8(1.3–2.5)	1.9(1.4–2.7)

NOTE. Values are expressed as n (%) unless otherwise indicated.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 2. 

  Kaplan–Meier plots displaying time to hypoxemia with (A) capnography blinded and open arms, subgrouped into (B) ERCP and EUS and (C) obesity and nonobesity.

Effect of Obesity on Hypoxemia and Apnea 

A total of 54 subjects (21.9%) had a body mass index ≥30 kg/m². There was evidence to suggest that obesity was an effect modifier of the capnography-hypoxemia association (P = .01) (Figure 2C). However, the significance was lost following adjustments for all factors used in multivariable HR for hypoxemia, as described in Table 3 (P = .13). Therefore, the interaction term was not included in the final model. Obesity was also an effect modifier of the capnography-apnea association (P = .03), and this interaction significance persisted even after adjustments (P = .041). There were no significant interactions between the use of diazepam and ASA class with capnography (P > .45).

Serious Adverse Events 

There were no serious adverse events during the trial such as death, respiratory failure, or use of reversal agents.

Back to Article Outline

Discussion 

We found that capnographic monitoring of respiratory activity during procedural sedation for elective ERCP and EUS significantly reduces hypoxemia, severe hypoxemia, and oxygen requirements by earlier detection and prompt correction of ventilation abnormalities. The benefit of capnography persisted even after oxygen supplementation. Thus, capnographic monitoring functions as an early warning system for hypoxemia and may further improve patient safety during gastrointestinal endoscopy.

Adverse events such as death, respiratory failure, or myocardial infarction occur rarely during endoscopy, with a reported frequency of 0.05%–0.2%.¹, ² Because an efficacy study will require thousands of patients with such rare events, we used hypoxemia as a surrogate marker due to its association with cardiac ischemia¹⁷, ¹⁸, ¹⁹ and post-ERCP pancreatitis²⁰ and its characterization as a sentinel critical event for subsequent adverse events.²¹ It is worth noting that the frequency of hypoxemia in our study was greater than expected from sample size estimations. However, it is at the higher end of the spectrum of the reported literature⁷, ¹¹, ¹², ¹³, ¹⁴, ¹⁵ and most likely due to lack of routine oxygen supplementation in accordance with our institutional practice. Even though oxygen use has been shown to reduce hypoxemic events during endoscopy,²² 2 large recent studies have revealed increased cardiopulmonary unplanned events with the routine use of supplemental oxygen.²³, ²⁴ This concept was physiologically proven in previous studies in which routine oxygen supplementation was found to mask apnea²⁵ and recurrent hypoxemia after oxygen supplementation was associated with respiratory arrest.⁷ However, ventilation monitoring along with oxygen supplementation has the potential to decrease hypoxemia to <20%, as shown in our study.

Previous attempts at capnographic monitoring with transcutaneous and side stream methods appeared promising but did not attain popularity due to technical failures.⁷, ⁹ However, the benefit of microstream capnography was shown in a pediatric study in which capnographic monitoring was found to significantly decrease hypoxemia during routine esophagogastroduodenoscopy and colonoscopy.¹⁰ During our study, we found that the capnography device performed well with high sensitivity. However, in 35 out of 263 patients (13%), the capnography erroneously displayed a flat line for at least 50 seconds without a concomitant decrease in oxygen saturation with normal chest excursions on subsequent clinical examination. We noticed that the most common reasons for this poor specificity (flat line with normal oxygen saturation) were obesity, narrow oropharyngeal inlet, and the use of wider-diameter scopes like the linear echoendoscope. We hypothesize that these false-positive results were probably caused by diminution of air stream due to narrow oropharyngeal inlet or blockage of the connecting tube with moisture. Even though the patients were able to maintain proper oxygenation, loss of stream was interpreted as no respiratory activity by the capnometer. Future technological innovations will likely overcome this fallacy.

Contrary to the prevailing notion that hypoxemia during procedural sedation occurs because of apnea or abnormal respiration, we found that up to one third of all hypoxemic events occurred with no change in ventilation pattern compared with baseline. The pathophysiological mechanisms of normal ventilation-associated hypoxemia should be explored in future studies, particularly if these findings are corroborated by other studies. In the meantime, we believe that the knowledge of normal ventilation during hypoxemic episodes will be of significant help to the endoscopist, particularly if additional medications were needed, which would otherwise be withheld under the current practice pattern for the fear of aggravating respiratory depression. Thus, capnographic monitoring may improve patient comfort by providing a physiologic basis for management of hypoxemia.

Our study has some limitations. First, even though hypoxemia and severe hypoxemia were assessed by 2 individuals independently, apnea was assessed by an independent observer with knowledge of the patient randomization allocation, which has the potential for introducing bias for assessment of apnea. Second, we enrolled patients undergoing elective ERCP/EUS and ASA class 1–3 with an opioid/benzodiazepine combination, and similar benefits may have to be confirmed for other endoscopic procedures, for ASA class 4–5 individuals, and for propofol sedation. Physiologic monitoring might be even more beneficial in sicker patients and those receiving propofol. Third, because this study was conducted at a single institution, the results have to be corroborated in other studies. However, we found no interaction between endoscopists and hypoxemia, suggesting that the results might be valid in other settings as well. Furthermore, the results are likely generalizable because the 7 endoscopists with different levels of experience probably represent the practice pattern of many centers. Fourth, we did not record the depth of sedation by any automated monitoring device during the study, although the endoscopists assessed the depth of sedation clinically as part of routine care. Even though it is conceivable that subjects with deeper levels of sedation have more hypoxemic episodes, and indeed those undergoing ERCP and EUS usually have deeper sedation,²⁶ we believe that randomization process would have led to equitable distribution of deep sedation episodes between the 2 groups. Thus, recording the depth of sedation would not have had any significant impact on the overall study results, but it is possible that knowledge of depth of sedation might have potentially improved our understanding of normal ventilation-hypoxemia, which may be caused by deep sedation.

In conclusion, capnographic monitoring of respiratory activity has the potential to improve patient safety during procedural sedation for ERCP/EUS due to the ventilation-based hypoxemia management paradigm. Simultaneous use of supplemental oxygen and capnography will likely decrease hypoxemic episodes significantly.

Back to Article Outline

Acknowledgments 

ClinicalTrials.gov no. NCT00675415.

The authors thank Edgar Achkar, MD, and Bret A. Lashner, MD, MPH, for offering valuable insight for this manuscript.

Back to Article Outline

References 

1. American Society for Gastrointestinal Endoscopy. Standards of Practice Committee (Sedation and anesthesia in GI endoscopy). Gastrointest Endosc. 2008;68:205–216
   • View In Article
   • Abstract
   • Full-Text PDF (557 KB)
   • CrossRef
2. American Society for Gastrointestinal Endoscopy. Guidelines for procedural sedation and monitoring during gastrointestinal endoscopy. Gastrointest Endosc. 2003;58:317–322
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (117 KB)
   • CrossRef
3. American Society of Anesthesiologists Task Force. Practice guidelines for sedation and analgesia by non-anesthesiologists. An updated report by the American Society of Anesthesiologists Task Force on Sedation and Analgesia by Non-Anesthesiologists. Anesthesiology. 2002;96:1004–1017
   • View In Article
   • MEDLINE
   • CrossRef
4. Kohn LT, Corrigan JM, Donaldson MS. To err is human, building a safer health system, Committee on Quality of Health Care in America. Washington, DC: Institute of Medicine, National Academy Press; 1999;
   • View In Article
5. The Joint Commission report. http://www.jointcommission.org/NR/rdonlyres/6530941D-98AD-4AC7-8944-9DDE1116E503/0/OBS_Standards_Sampler_2007_final.pdfAccessed July 9, 2008
   • View In Article
6. Vargo JJ, Zuccaro G, Dumot JA, et al. Automated assessment of respiratory activity is superior to pulse oximetry and visual assessment for the detection of early respiratory depression during therapeutic upper endoscopy. Gastrointest Endosc. 2002;55:826–831
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (112 KB)
   • CrossRef
7. Luft KF. Uber eine neue methode der registrierenden gasanalyse mit hiffe der absorption ultraroter strahlen ohne spektrale zerlegung. Zeitschrift Fur Technische Physik. 1943;24:97–100
   • View In Article
8. Freeman ML, Hennessy JT, Cass OW, et al. Carbon dioxide retention and oxygen retention during gastrointestinal endoscopy. Gastroenterology. 1993;105:331–339
   • View In Article
   • Abstract
   • Full-Text PDF (1056 KB)
9. Colman Y, Krauss B. Microstream capnography technology: a new approach to an old problem. J Clin Monit Comput. 1999;15:403–409
   • View In Article
   • MEDLINE
   • CrossRef
10. Lightdale JR, Goldmann DA, Feldman HA, et al. Microstream capnography improves patient monitoring during moderate sedation: a randomized, controlled trial. Pediatrics. 2006;117:1170–1178
    • View In Article
11. Vargo JJ, Zuccaro G, Dumot JA, et al. Gastroenterologist-administered propofol versus meperidine and midazolam for advanced upper endoscopy: a prospective, randomized trial. Gastroenterology. 2002;123:8–16
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (97 KB)
    • CrossRef
12. Woods SDS, Chung SCS, Leung JWC, et al. Hypoxia and tachycardia during endoscopic retrograde cholangiopancreatography: detection by pulse oximetry. Gastrointest Endosc. 1989;35:523–525
    • View In Article
    • Abstract
    • Full-Text PDF (1968 KB)
    • CrossRef
13. Wong PYN, Lane MR, Hamilton I. Arterial oxygen desaturation during endoscopic retrograde cholangiopancreatography. Endoscopy. 1993;25:309–310
    • View In Article
    • MEDLINE
    • CrossRef
14. Griffin MS, Chung SCS, Leung JWC, et al. Effects of oxygen and hypoxemia and tachycardia during endoscopic retrograde cholangiopancreatography. Br Med J. 1990;300:83–84
    • View In Article
15. Haines DJ, Bibbey J, Green JRB. Does nasal oxygen reduce the cardiorespiratory problems experienced by elderly patients undergoing endoscopic retrograde cholangiopancreatography?. Gut. 1992;33:973–975
    • View In Article
    • MEDLINE
    • CrossRef
16. Qadeer MA, Vargo JJ, Dumot JA, et al. Capnography prevents hypoxemia in patients undergoing ERCP and EUS (A randomized controlled trial (interim analysis)). (abstr) Gastrointest Endosc. 2008;67:AB84
    • View In Article
    • Full Text
    • Full-Text PDF (45 KB)
    • CrossRef
17. Holm C, Christensen M, Rasmussen V, et al. Hypoxaemia and myocardial ischaemia during colonoscopy. Scand J Gastroenterol. 1998;33:769–772
    • View In Article
    • MEDLINE
    • CrossRef
18. Johnston SD, McKenna A, Tham TC. Silent myocardial ischaemia during endoscopic retrograde cholangiopancreatography. Endoscopy. 2003;35:1039–1042
    • View In Article
    • MEDLINE
    • CrossRef
19. Jurell KR, O'Connor KW, Slack J, et al. Effect of supplemental oxygen on cardiopulmonary changes during gastrointestinal endoscopy. Gastrointest Endosc. 1994;40:665–670
    • View In Article
    • Abstract
    • Full Text
20. Fisher L, Fisher A, Thomson A. Cardiopulmonary complications of ERCP in older patients. Gastrointest Endosc. 2006;63:948–955
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (136 KB)
    • CrossRef
21. Coté CJ, Notterman DA, Karl HW, et al. Adverse sedation events in pediatrics: a critical incident analysis of contributing factors. Pediatrics. 2000;105:805–814
    • View In Article
    • MEDLINE
    • CrossRef
22. Wang CY, Ling LC, Cardosa MS, et al. Hypoxia during upper gastrointestinal endoscopy with and without sedation and the effect of pre-oxygenation on oxygen saturation. Anaesthesia. 2000;55:654–658
    • View In Article
    • MEDLINE
    • CrossRef
23. Sharma VK, Nguyen CC, Crowell MD, et al. A national study of cardiopulmonary unplanned events after GI endoscopy. Gastrointest Endosc. 2007;66:27–34
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (778 KB)
    • CrossRef
24. Cohen LB, Holub J, Lieberman DA, et al. Does routine use of supplemental oxygen during endoscopy really reduce the risk of cardiopulmonary complications?. Gastrointest Endosc. 2007;65:AB103
    • View In Article
    • Full Text
    • Full-Text PDF (47 KB)
    • CrossRef
25. Zuccaro G, Radaelli F, Vargo J, et al. Routine use of supplemental oxygen prevents recognition of prolonged apnea during endoscopy. (abstr) Gastrointest Endosc. 2000;51:AB141
    • View In Article
26. Qadeer MA, Vargo JJ, Patel S, et al. Bispectral index monitoring of conscious sedation with the combination of meperidine and midazolam during endoscopy. Clin Gastroenterol Hepatol. 2008;6:102–108
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (210 KB)
    • CrossRef