Dig Dis Sci (2009) 54:1035–1040 DOI 10.1007/s10620-008-0452-2
ORIGINAL ARTICLE
Risk Factors for Hypoxemia During Ambulatory Gastrointestinal Endoscopy in ASA I–II Patients Mohammed A. Qadeer Æ A. Rocio Lopez Æ John A. Dumot Æ John J. Vargo
Received: 1 June 2008 / Accepted: 16 July 2008 / Published online: 12 November 2008 Ó Springer Science+Business Media, LLC 2008
Abstract Background Most studies identify the American Society of Anesthesiology (ASA) classification as the most significant risk factor for hypoxemia. The risk factors operative within ASA I and II patients are not well defined. Therefore, we analyzed prospectively collected data to identify the risk factors of hypoxemia in such patients. Methods A combination of a narcotic and benzodiazepine was used for sedation and oxygen was supplemented if hypoxemia (oxygen saturation B90%) developed. Univariate and multivariate analyses were performed and correlations estimated for predetermined clinical variables. Results 40 of 79 patients (51%) developed hypoxemia, which occurred more frequently in the obese (71%; 10/14) than the nonobese (46%; 30/65) group (P = 0.08). On multivariate analysis, the odds ratios (OR) and 95% confidence intervals (CI) for developing hypoxemia were age C 60 years 4.5 (1.4–14.3) P = 0.01, and incremental 25-mg doses of meperidine 2.6 (1.02–6.6) P = 0.04. Body mass index (BMI) significantly correlated with the number of hypoxemic episodes (rho 0.26, 95% CI 0.04–0.48, P = 0.02). Conclusion In ASA I and II patients, BMI significantly correlated with the number of hypoxemic episodes, whereas age C 60 years and meperidine dose were significant risk factors for hypoxemia.
M. A. Qadeer J. A. Dumot J. J. Vargo (&) Department of Gastroenterology and Hepatology, Digestive Disease Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue, A 30, Cleveland, OH 44195, USA e-mail:
[email protected] A. Rocio Lopez Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, USA
Keywords Hypoxemia Ambulatory endoscopy ASA I and II Body mass index
Introduction In the United States, the majority of gastrointestinal endoscopic procedures are performed under conscious or moderate sedation in the ambulatory setting [1–3]. Conscious sedation reduces patient discomfort by inducing sedation, analgesia and amnesia during the procedure and facilitates endoscopy due to better patient tolerance [1–3]. However, conscious sedation is also the cause of many potentially serious unplanned events during endoscopy. The most common unplanned cardiopulmonary event is hypoxemia, which occurs in up to 10–70% of patients depending on the definition of hypoxemia, the patient population, the level of sedation, and the type of endoscopy [4–9]. Even though most cases of hypoxemia are transient and respond to oxygen supplementation, it has been shown to be associated with cardiac ischemia in the form of STsegment depression [10, 11] in people with or without preexisting cardiac disease. Hypoxemia may also result in unsatisfactory sedation from the patients’ perspective due to the fear of inducing further hypoxemia and delay or cancellation of the procedure. Therefore, avoiding hypoxemia, as well as its prompt identification and treatment, is important to prevent ensuing serious complications during endoscopy. Previous studies have evaluated the risk factors of hypoxemia in both retrospective and prospective methods with disparate results [4–13]. A majority of the reports found the American Society of Anesthesiology (ASA) classification to be the most significant overall risk factor. Due to the confounding effect of the ASA class within the
123
1036
diverse populations in the previous reports, risk factors of hypoxemia within ASA I and II patients, who constitute the majority of ambulatory patients, are not well defined. Secondly, the relationships between body weight and hypoxemia have not been clearly defined in previous studies [8, 12, 13]. Therefore, our aims for this study were to evaluate the risk factors of hypoxemia in ASA I and II patients during moderate sedation for ambulatory endoscopic procedures and to evaluate the relationship between body mass index (BMI) and hypoxemia.
Dig Dis Sci (2009) 54:1035–1040
episodes with misplaced sensor were not considered hypoxemic. We also recorded all episodes of repeated hypoxemia (oxygen saturation B90%) after the first episode. Normal saline was given if the systolic blood pressure dropped below 90 mmHg. At the end of procedure, patients were transferred to a recovery area, where they were asked to rate the degree of comfort on a visual analog scale (VAS) after attaining full consciousness (MOAA/S level 5). VAS variables included patient satisfaction (0 = very dissatisfied to 100 = very satisfied), overall discomfort (0 = none to 100 = severe) and gagging (0 = none to 100 = severe).
Methods Sedative Dosages Eligibility Criteria The study was reviewed and approved by the institutional review board of the Cleveland Clinic Foundation (Cleveland, OH, USA). Consecutive patients older than 18 years with ASA class I and II who were able to give informed consent for elective ambulatory endoscopy were offered participation. Patients were excluded if they were unable or unwilling to provide informed consent, undergoing emergency procedures, had allergy to study medications, were pregnant, or belonged to ASA class III–V. Gastroenterologists with an experience of at least 5,000 endoscopies performed the procedures. Patient Assessment and Data Collection A focused medical history, physical examination, and questionnaire were completed in the preprocedure area to record the following baseline data: age, gender, height, weight, use of sedative and narcotic medicines, and history of smoking and alcohol intake. We defined obesity as BMI C 30 kg/m2 and nonobese as BMI \ 30 kg/m2. All patients were monitored with continuous pulse oximetry on room air, blood pressure, and heart-rate measurements (Passport 2, Datascope Corp, Paramus, NJ, USA). An independent observer recorded the patients’ blood pressure, heart rate, oxygen saturation, and medication timing and dose. The observer also recorded cardiopulmonary events defined as oxygen desaturation \90%, hypotension of \90 mmHg systolic and 50 mmHg diastolic blood pressures, and need for reversal of agents or respiratory assistance with airway interventions. Patients received oxygen through nasal cannula if the oxygen saturation dropped to \90% anytime during the procedure in concordance with our institutional practice pattern. We defined hypoxemia as oxygen saturation of B90% anytime during the procedure, regardless of episode duration. Before labeling any oxygen desaturation as hypoxemia, we checked whether the pulse oximetry sensor was properly attached, and all desaturation
123
The endoscopist performing the procedures determined sedative and analgesic doses according to the protocol described in our previous study [14]. Initial intravenous doses of meperidine at 50–75 mg and midazolam at 2–3 mg were given based on patient age and weight. Additional doses were titrated based on endoscopist and nurse assessment of patient tolerance. Increments of meperidine (25–50 mg) and midazolam (1–2 mg) were given if patients showed signs of discomfort, restlessness, or agitation that were not related to hypoxemia. Statistical Methods Descriptive statistics were computed for all variables. These included medians and 25th and 75th percentiles for continuous factors and frequencies for categorical variables. Wilcoxon rank sum tests were used to assess association between several continuous factors, such as BMI and hypoxemia. Hypotension was not included in the model due to very few observations (n = 7). Pearson’s chi-square or Fisher’s exact tests were used for categorical factors. In addition, multivariate logistic regression analysis was performed. Final multivariate models were selected using a stepwise selection method, with 0.50 and 0.25 as the significance criterions for a variable entering and remaining in the model, respectively. We estimated the association between meperidine and midazolam velocity and the incidence of hypoxemia. A P \ 0.05 was considered statistically significant. SAS version 9.1 software (SAS Institute, Cary, NC, USA) was used to perform all analyses.
Results Descriptive Characteristics A total of 80 consecutive patients were offered enrollment, and all of them agreed to participate. One patient was
Dig Dis Sci (2009) 54:1035–1040
1037
excluded due to discrepancy in the demographic information and BMI calculation. Thus, the final analysis was performed on 79 patients. Table 1 lists the descriptive characteristics of the study cohort. Hypoxemia developed in 51% (40/79) of patients at least once during the procedure, and 20% of the cohort had more than one episode of hypoxemia. Table 1 Overall descriptive statistics (n = 79) Factor Age
Female Body mass index (BMI) (kg/m2) Obese (BMI C 30) Total meperidine dose (mg) Total midazolam dose (mg)
Hypoxemia was more common in patients C60 years compared with those \60 years (68% vs. 44%, P = 0.032) (Table 2). The median BMI of patients developing hypoxemia was higher than the median BMI of patients without hypoxemia (26.9 vs. 25.0, respectively P = 0.04). Similarly, hypoxemia was numerically more common in obese patients compared with nonobese patients 71%
Statisticsa 62.0 (50.0, 74.0)
Gender Male
Univariate Analyses
Table 2 Possible predictors of hypoxemia: univariate analysis Factor
42 (53.2) 37 (46.8) 25.7 (22.9, 29.1) 14 (17.7)
20 (25.3)
Colonoscopy
20 (25.3)
13 (32.5)
22 (56.4)
Yes
27 (67.5)
17 (43.6)
23 (57.5) 17 (42.5)
19 (48.7) 20 (51.3)
Male Female BMI (kg/m2)
P value 0.032
No Gender
Procedure EGD
No hypoxemiaa (n = 39)
Age C 60 years
100.0 (75.0, 100.0) 4.0 (3.0, 6.0)
Hypoxemiaa (n = 40)
0.43
26.9 (23.5, 29.7)
25.0 (21.5, 26.7)
10/14 (71%)
19 (24.1)
Total meperidine dose (mg)
100.0 (75.0, 100.0) 100.0 (75.0, 100.0) 0.88
1
45 (57.0)
Total midazolam dose (mg)
2
11 (13.9)
Procedure
3
4 (5.1)
4
19 (24.1)
Alcohol Yes
10 (12.7)
ERCP EUS
20 (25.3)
Procedure per physician
No
EGD Colonoscopy ERCP
69 (87.3)
EUS
4.0 (3.5, 5.0)
30/65 (46%)
0.04
Obesity
4.0 (3.0, 6.0)
9 (22.5) 12 (30.0)
11 (28.2) 8 (20.5)
8 (20.0)
12 (30.8)
11 (27.5)
8 (20.5) 0.063
1
26 (65.0)
19 (48.7)
Yes
11 (13.9)
2
7 (17.5)
4 (10.3)
No
68 (86.1)
3
0 (0.0)
4 (10.3)
4
7 (17.5)
12 (30.8)
Use of narcotics Yes
12 (15.2)
Alcohol
No
67 (84.8)
Yes
6 (15.0)
4 (10.3)
No
34 (85.0)
35 (89.7)
Yes
20 (25.3)
Tobacco
No
59 (74.7)
Yes
4 (10.0)
7 (18.0)
No
36 (90.0)
32 (82.1)
Use of sedatives/antidepressants
Number of times hypoxemia observed during procedure
0.74
0.31
0
39 (49.4)
1
23 (29.1)
2
12 (15.2)
3 4
1 (1.3) 1 (1.3)
5
2 (2.5)
Yes
11 (27.5)
9 (23.1)
6
1 (1.3)
No
29 (72.5)
30 (76.9)
Use of narcotics Yes No
0.99 0.52
Physician
Tobacco
0.08
0.054 3 (7.5)
9 (23.1)
37 (92.5)
30 (76.9)
Use of sedatives/ antidepressants
0.65
EGD esophagogastroduodenoscopy, ERCP retrograde cholangiopancreatography, EUS endoscopic ultrasound
BMI body mass index, EGD esophagogastroduodenoscopy, ERCP retrograde cholangiopancreatography, EUS endoscopic ultrasound
a
a
Values presented are median (Q25, Q75) or n (%)
Values presented are median (Q25, Q75) or n (%)
123
1038
Dig Dis Sci (2009) 54:1035–1040
(10/14) vs. 46% (30/65), respectively P = 0.08. There were no significant differences in the dosages of medications or alcohol or tobacco use in the two groups. Hypoxemia was less common in people using narcotic medications 8% vs. 23% P = 0.054. Multivariate Analyses In the final multivariate model that included age (C60 or \60 years), BMI, total meperidine dose, total midazolam dose, and use of narcotics, the significant factors that predicted hypoxemia were age C 60 years and total meperidine dose (Table 3). Patients C60 years had four times the risk of developing hypoxemia compared with \60 years (OR 4.1, 95% CI 1.3–13.0, P = 0.02). With every 25-mg increment in meperidine dose, the risk of hypoxemia increased by 2.5 times (OR 2.5, 95% CI 1.04– 6.2). The effect of BMI for predicting hypoxemia was not statistically significant, with every 5-U increase in the BMI, the risk of hypoxemia increased by 1.6 times (OR 1.6, 95% CI 0.95–2.6, P = 0.079).
Table 4 Spearman’s correlation coefficients examining the relationship between body mass index (BMI), number of hypoxemic events, and total meperidine and midazolam dosage Factors
Rho
95% CI
P value
Total meperidine dose (mg)
0.06
(-0.28, 0.17)
0.61
Total midazolam dose (mg) Number of hypoxemia events
0.05 0.26
(-0.28, 0.17) (0.04, 0.48)
0.65 0.022
Total meperidine dose (mg)
0.09
(-0.14, 0.32)
0.43
Total midazolam dose (mg)
0.07
(-0.16, 0.29)
0.56
BMI
Number of hypoxemia events
Table 5 No significant differences were found in patient satisfaction scores based on visual analog scale (VAS) scores between those who developed hypoxemia and those who did not Factor
Hypoxemia (n = 40)
No hypoxemia (n = 39)
P value
Discomfort
7.0 (1.0, 18.0)
9.0 (2.0, 17.0)
0.64
Gagging
6.0 (0.0, 9.0)
5.0 (0.0, 14.0)
0.89
Sedation satisfaction 97.5 (92.0, 100.0) 96.0 (89.0, 100.0) 0.54
Correlation Between Hypoxemia, BMI, and Meperidine and Midazolam Dosages Discussion There was a significant correlation between the number of hypoxemic episodes and BMI (rho 0.26, 95% CI 0.04–0.48, P = 0.02) (Table 4). No such correlations were found between total meperidine and midazolam dosages and BMI or the number of hypoxemic episodes and total meperidine and midazolam dosages. Effect of Hypoxemia on Patient Satisfaction Scores There were no significant differences in patient satisfaction scores between the group developing hypoxemia and that with no hypoxemia (Table 5).
Table 3 Possible predictors of hypoxemia: multivariate analysis Factor
Reference
OR (95% CI)
Age C 60 years
Yes vs. no
4.1 (1.3, 13.0)
0.02
BMI
5 kg/m2
1.6 (0.95, 2.6)
0.079
Total meperidine dose (mg)
25 mg
2.5 (1.04, 6.2)
0.041
Total midazolam dose (mg)
1 mg
0.74 (0.46, 1.2)
Use of narcotics
No versus yes
3.2 (0.64, 15.6)
P value
0.2 0.16
BMI body mass index, OR odds ratio, CI confidence intervals
123
Our results suggest that hypoxemia occurs in about 50% of ASA I and II patients undergoing ambulatory gastrointestinal endoscopy. Age C 60 years and total meperidine dose are significant predictors of hypoxemia. We also found BMI correlated with the number of hypoxemic episodes and that hypoxemia was numerically more frequent in obese patients compared with nonobese patients. Risk factors for hypoxemia have been evaluated in several studies [4–13], but a lack of a uniform definition of hypoxemia makes comparison across the studies difficult. This difficulty is further compounded by differences in patient populations, type of endoscopy, use of supplemental oxygen, and type of sedation agents [4–15]. Therefore, the incidence of hypoxemia during conscious sedation for gastrointestinal endoscopy has ranged widely, between 10% and 70% [4–15]. Overall, conscious sedation is overwhelmingly safe, with a very low risk of death, requirement for artificial ventilation, myocardial infarction, or central nervous system damage. Therefore, hypoxemia is used as a surrogate for these outcomes due to rarity of these hard events and also because hypoxemia is a precursor to these outcomes [16–19]. As the majority of studies included heterogeneous patient groups, further risk factor assessment in these studies has been potentially confounded by the strong influence of ASA class. Therefore, specific risk factors in ASA class I and II patients are
Dig Dis Sci (2009) 54:1035–1040
not entirely clear. Thus, our study sheds light on the risk factors of hypoxemia in a cohort of such patients. There are several possible mechanisms for hypoxemia during conscious sedation [20]. Most common mechanisms include hypoventilation due to depression of the respiratory drive by sedation and analgesic medications. In addition, the process of esophageal intubation causes voluntary transient apnea [21] or some degree of ventilation-perfusion mismatch. Therefore, sedation and intubation appear to be the strongest risk factors of apnea, which may in turn lead to hypoxemia. Longer duration of procedures has also been associated with increased incidence of hypoxemia [22]. Therapeutic procedures, due to their longer duration, are therefore more likely to be associated with hypoxemia. Body weight has been weakly correlated with hypoxemia in at least three studies [8, 12, 13]. However, all those studies suffered from a low sample of obese patients and unclear definitions of hypoxemia and BMI, which led to disparate results. Thus, the results of our study may help elucidate the correlation between body weight and hypoxemia a little better. The fact that obesity is correlated with the number of hypoxemic episodes (Table 4) despite oxygen supplementation suggests either a mechanical problem related to body weight and airway tissues, or other unknown mechanisms associated with obesity such as sleep apnea (subclinical and/or undiagnosed). Increasing age has been related to hypoxemia in several studies [5, 23] which may be due to decreasing body fat stores with resultant decrease in the reservoir capacity for binding lipophilic sedative and narcotic agents. Thus, smaller doses of these agents will be more active biologically due to increased bioavailability. Alterations in hepatic metabolism of sedatives due to decreases in perfusion, protein binding, or competition for metabolic pathways by other medications may also play a role. Additionally, elderly people might also have alterations in ventilation-perfusion matching, making them more prone to hypoxemia. Our study has some inherent weaknesses. First, the sample size was small, which may result in some false negative results (type II phenomenon), which may explain the nonsignificant difference in hypoxemia between the various types of procedures. However, our sample size was estimated for assessing accuracy of bispectral index monitoring of conscious sedation during endoscopy, which was the main result of this study [24]. This study was performed with four different gastroenterologists, and it is conceivable that patients of these four gastroenterologists have different rates of hypoxemia. However, we were unable to find any significant differences among them. Our study design represents the practice pattern in a diverse group of gastroenterologists, and hence our results may be generalizable. Due to the small number of obese people in the cohort, we were not able to find a statistically
1039
significant difference between obese and nonobese groups, and BMI lost its significance in the multivariate analysis, although it does appear that obesity plays a significant role in causing hypoxemia. In conclusion, hypoxemia during ambulatory gastrointestinal endoscopy is related to age and meperidine dose. People with higher BMI tend to have recurrent hypoxemic episodes despite oxygen use, implying that the mechanism might be mechanical obstruction or some other unknown cause. Therefore, we recommend that obese patients, and particularly those [60 years, should be monitored closely during conscious sedation.
References 1. ASGE Standards of Practice Committee. Sedation and monitoring of patients undergoing gastrointestinal endoscopic procedures. Gastrointest Endosc. 1995;42:626–629. doi:10.1016/S0016-5107 (95)70031-5. 2. American Society for Gastrointestinal Endoscopy. Guidelines for conscious sedation and monitoring during gastrointestinal endoscopy. Gastrointest Endosc. 2003;58:317–322. doi:10.1067/ S0016-5107(03)00001-4. 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. doi:10.1097/00000542-200 204000-00031. 4. Vargo JJ, Holub JL, Faigel DO, Lieberman DA, Eisen GM. Risk factors for cardiopulmonary events during propofol-mediated upper endoscopy and colonoscopy. Aliment Pharmacol Ther. 2006;24:955–963. doi:10.1111/j.1365-2036.2006.03099.x. 5. Sharma VK, Nguyen CC, Crowell MD, Lieberman DA, de Garmo P, Fleischer DE. A national study of cardiopulmonary unplanned events after GI endoscopy. Gastrointest Endosc. 2007;66:27–34. doi:10.1016/j.gie.2006.12.040. 6. Iber FL, Sutberry M, Gupta R, Kruss D. Evaluation of complications during and after conscious sedation for endoscopy using pulse oximetry. Gastrointest Endosc. 1993;39:620–625. 7. Fisher L, Fisher A, Thomson A. Cardiopulmonary complications of ERCP in older patients. Gastrointest Endosc. 2006;63:948– 955. doi:10.1016/j.gie.2005.09.020. 8. Woods SD, Chung SC, Leung JW, Chan AC, Li AK. Hypoxia and tachycardia during endoscopic retrograde cholangiopancreatography: detection by pulse oximetry. Gastrointest Endosc. 1989;35:523–525. 9. Wong PYN, Lane MR, Hamilton I. Arterial oxygen desaturation during endoscopic retrograde cholangiopancreatography. Endoscopy. 1993;25:309–310. 10. Holm C, Christensen M, Rasmussen V, Schulze S, Rosenberg J. Hypoxaemia and myocardial ischaemia during colonoscopy. Scand J Gastroenterol. 1998;33:769–772. doi:10.1080/0036552 9850171747. 11. Johnston SD, McKenna A, Tham TC. Silent myocardial ischaemia during endoscopic retrograde cholangiopancreatography. Endoscopy. 2003;35:1039–1042. doi:10.1055/s-2003-44597. 12. Block R, Jankowski J, Johnston D, Colvin JR, Wormsley KG. The administration of supplementary oxygen to prevent hypoxia during upper alimentary endoscopy. Endoscopy. 1993;25:269–273.
123
1040 13. Dhariwal A, Plevris JN, Lo NT, Finlayson ND, Heading RC, Hayes PC. Age, anemia, and obesity-associated oxygen desaturation during upper gastrointestinal endoscopy. Gastrointest Endosc. 1992;38:684–688. 14. Vargo JJ, Zuccaro G, Dumot JA, et al. Gastroenterologistadministered propofol versus meperidine and midazolam for advanced upper endoscopy: A prospective, randomized trial. Gastroenterology. 2002;123:8–16. doi:10.1053/gast.2002.34232. 15. Iwao T, Toyonaga A, Harada H, et al. Arterial oxygen desaturation during non-sedated diagnostic upper gastrointestinal endoscopy in patients with cirrhosis. Gastrointest Endosc. 1994; 40:281–284. 16. Schenck J, Mu¨ller CH, Lu¨bbers H, Mahlke R, Lehnick D, Lankisch PG. Does gastroscopy induce myocardial ischemia in patients with coronary heart disease? Endoscopy. 2000;32:373– 376. doi:10.1055/s-2000-9005. 17. Seinela¨ L, Reinikainen P, Ahvenainen J. Effect of upper gastrointestinal endoscopy on cardiopulmonary changes in very old patients. Arch Gerontol Geriatr. 2003;37:25–32. doi:10.1016/ S0167-4943(03)00002-5. 18. Haines DJ, Bibbey J, Green JR. Does nasal oxygen reduce the cardiorespiratory problems experienced by elderly patients undergoing endoscopic retrograde cholangiopancreatography? Gut. 1992;33:973–975. doi:10.1136/gut.33.7.973. 19. Griffin MS, Chung SCS, Leung JWC, et al. Effects of oxygen and hypoxemia and tachycardia during endoscopic retrograde cholangiopancreatography. BMJ. 1990;300:83–84.
123
Dig Dis Sci (2009) 54:1035–1040 20. Rimmer KP, Graham K, Whitelaw WA, Field SK. Mechanisms of hypoxemia during panendoscopy. J Clin Gastroenterol. 1989;11:17–22. doi:10.1097/00004836-198902000-00006. 21. Vargo JJ, Zuccaro G, Dumot JA, Conwell DL, Morrow JB, Shay SS. 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. doi:10.1067/mge.2002. 124208. 22. Mu¨ller S, Prolla JC, Maguilnik I, Breyer HP. Predictive factors of oxygen desaturation of patients submitted to endoscopic retrograde cholangiopancreatography under conscious sedation. Arq Gastroenterol. 2004;4:162–166. 23. Gangi S, Saidi F, Patel K, et al. Cardiovascular complications after GI endoscopy: occurrence and risks in a large hospital system. Gastrointest Endosc. 2004;60:679–685. doi:10.1016/ S0016-5107(04)02016-4. 24. Qadeer MA, Vargo JJ, Patel S, et al. Utility of bispectral index monitoring of conscious sedation with the combination of meperidine and midazolam during gastrointestinal endoscopy. Clin Gastroenterol Hepatol. 2008;6:102–108. doi:10.1016/j.cgh. 2007.10.005.
