Abstract
Background & Aims:
Despite the increased use of anesthesia services for endoscopic procedures in the United States, the risks of anesthesia-directed sedation (ADS) are unclear. We analyzed national data from multiple centers to determine patterns of use of anesthesia services and risk factors for serious complications.
Methods:
We performed a cross-sectional study using the National Anesthesia Clinical Outcomes Registry, a national quality improvement database. Univariable and bivariate analyses investigated frequencies and relationships between pre-defined variables and serious complications of anesthesia (cardiovascular, respiratory, neurologic, drug-related, patient injury, death, or unexpected admission). A multivariable mixed effects model determined odds ratios between these variables and serious complications, adjusting for confounders and varying reporting practices.
Results:
In total, 428,947 endoscopic procedures of adults were performed using ADS from 2010 from 2015. The population was 54.9% female with a mean age of 59.1 years and predominantly American Society of Anesthesiologists class 2/3 (74.4%). More than half of the procedures were colonoscopies (51.4%); 37.4% were esophagogastroduodenoscopies and 6.5% were endoscopic retrograde cholangiopancreatographies. A total of 4441 complications (1.09%) were reported; 1349 were serious complications (0.34%). In multivariable analysis, older age, American Society of Anesthesiologists class 4/5, esophagogastroduodenoscopy, general anesthesia, endoscopists on an overnight shift, and longer cases were independently and significantly associated with serious complications.
Conclusions:
In an analysis of data from the National Anesthesia Clinical Outcomes Registry, we found ADS during endoscopy to be safe, with few serious complications (<1% of procedures). Risk of ADS complications increased with older age, more severe disease, procedure type, and case complexity.
Keywords: ASA, EGD, ERCP, colonoscopy
Background
Anesthesia services for gastrointestinal (GI) endoscopic procedures have risen dramatically over the years.1,2 Prior to 2004, only a fraction of endoscopic procedures were performed with anesthesia assistance.2,3 By 2009, 30–35% of upper endoscopies and colonoscopies had anesthesia directed sedation (ADS),2,4 and this number increased to half of colonoscopies by 2012.5 Despite the steady increase in anesthesia services and Propofol use in endoscopic procedures, it remains unclear whether ADS is associated with increased complications as compared to endoscopist-directed moderate sedation. Some studies have suggested increased risk of aspiration pneumonia,6,7 colonic perforation, bleeding, and cardiovascular events with ADS.8 Whereas other studies have shown no increased risk of complication after accounting for disease severity and case complexity, which may be driving the utilization of anesthesia services and its increased complications.9–12 Because sicker patients are more likely to require anesthesiologist support, the extant literature is conflicting regarding the true risks associated with deep sedation during endoscopic procedures.
Given the evolving role of ADS for gastrointestinal procedures and the inconsistent findings of prior research, we conducted a large epidemiological study to determine the risk of complications associated with ADS during GI endoscopic procedures. We hypothesized that certain patient, facility, and procedural factors may increase risk of complications. Given conflicting evidence as to whether esophagogastroduodenoscopy (EGD)10 or colonoscopy13 is associated with more anesthetic complications, we also aimed to determine the risk profile for each procedure type.
Methods
Database and Population
This cross-sectional study utilized the National Anesthesia Clinical Outcomes Registry (NACOR), a national registry comprised of inpatient and outpatient anesthesia procedure data conducted at over 2700 facilities across the U.S. since January 1st 2010. NACOR comprises the largest anesthesia database in the U.S. sponsored by the Anesthesia Quality Institute (AQI). Over 150 variables are reported including age, sex, International Classification of Diseases (ICD) 9 codes, American Society of Anestheisa (ASA) classification, and procedure type as defined by surgical and anesthesia current procedural terminology (CPT) codes. Data is automatically imported from each clinical site’s electronic medical record and is audited regularly by the AQI, including several levels of data validation to ensure maximum data integrity. Outcomes data including complications are available if reported by the clinical site. NACOR has been used previously for other epidemiological studies of anesthesia utilization, complications, and quality.14,15 Pediatric cases (<18 years) were excluded, as well as gastrointestinal surgical cases performed in the operating room. This study received exemption from the University of North Carolina Institutional Review Board due to its use of deidentified data.
Variables and Outcomes of Interest
For all endoscopic procedures, we investigated: patient characteristics (age, sex, ASA classification), facility characteristics (U.S. region, hospital/center type, facility volume), procedure type (EGD/enteroscopy/endoscopic ultrasound (EUS), colonoscopy/flexible sigmoidoscopy, and endoscopic retrograde cholangiopancreatography (ERCP)), other procedural characteristics (day/night shift, emergent/inpatient status, weekday/holiday status), and anesthesia characteristics (sedation type, case duration). If more than one sedation type was listed, general anesthesia took precedence and was reported as the primary anesthesia type. Facility volume was defined as the number of concurrent cases taking place with anesthesia services at the time of procedure, and was categorized as low (0–9), medium (10–27), and high (>27) volume. Case duration was defined as anesthesia start to finish time (and therefore is generally longer and not equal to endoscopy procedure duration). Case duration was categorized as short (1–26 minutes), medium (27–40 minutes), and long (>40 minutes).
A total of 47 anesthetic complication outcomes are recorded by an anesthesia provider at the end of a case and reported in NACOR (Table 1). Because NACOR primarily collects data on anesthesia complications, no specific GI or endoscopic procedural outcomes are recorded in this database. Our primary outcome of interest was a composite outcome of any serious complication defined as a respiratory, cardiovascular, drug, neurologic, patient injury, unplanned admission or death complication.
Table 1.
Definition of Serious Anesthetic Complication During Endoscopic Procedures
| COMPLICATION CATEGORY | EXAMPLES |
|---|---|
| CARDIOVASCULAR | Arrhythmia Cardiac arrest Hemodynamic instability Hypotension Myocardial infarction Myocardial ischemia Transfusion |
|
| |
| DEATH | Death due to any cause |
|
| |
| DRUG | Adverse reaction Anaphylaxis Malignant hyperthermia Medication error |
|
| |
| NEUROLOGIC | Awareness Emergence Neurologic deficit Seizure Stroke Vision loss |
|
| |
| PATIENT INJURY | Eye injury Infection Other bodily injury |
|
| |
| RESPIRATORY | Airway obstruction Aspiration Difficult airway Pneumothorax Pulmonary embolus Reintubation Respiratory arrest |
|
| |
| UNPLANNED ADMISSION | Admission to hospital or intensive care unit |
Statistical Analysis
We examined means and standard deviations for each continuous variable, and frequencies for each categorical variable. Missing data were examined for each variable; those with a large number of missing observations were excluded from the final model (e.g. outpatient/inpatient status, weekday/weekend). For certain variables indicating rare events that were likely to be reported with high specificity, missing data were imputed as negative. For example, cases not classified as ‘emergent’ were assumed to be non-emergent or elective cases, as based on the ASA classification system.
We performed bivariate comparisons of each variable with the dichotomous outcome of serious complication using Pearson’s chi-square tests for categorical predictors, and t-tests or individual logistic regression models for continuous predictors. We also examined relationships among the predictor variables for any instances of collinearity that could affect the validity of modeling estimates. Frequency of complications was calculated by dividing the number of complications by the total number of procedures with complete complication data reported (defined as the total cases at risk for complication).
We fit a logistic regression model to determine risk factors that best predicted the probability of developing a serious complication reported as odds ratios. Our model comprised a hierarchical mixed-effects model adjusted for clustering by facility, practice and provider to account for variability in reporting of complications. More specifically, an identification code for practice, facility, and provider was used in the final multivariable model to account for potential reporting bias. Our final multivariable model was restricted to observations that had non-missing data (Figure 1). The final model consisted of variables that were reported to be associated with complications in the literature (e.g. procedure type, ASA, sedation type), potential predictors of complications that could be clinically relevant (e.g. age, sex, emergent status), as well as those variables that were statistically significant on bivariate analysis. We also performed subgroup analyses to investigate complications stratified by procedure type. Two-tailed p values are reported and p < 0.05 was considered statistically significant. Stata 15 (Stata Corporation, College Station, TX, USA) was used for all analyses.
Figure 1.
Flow chart depicting cases in the NACOR database and the study population.
Results
Patient and Facility Characteristics
Our sample consisted of 476,153 endoscopic procedures with anesthesia services. Among these, 428,947 procedures met our inclusion criteria (Figure 1). The top 5 ICD9 codes associated with these endoscopic procedures are seen in Supplemental Table 1. On univariate analysis, more than half of the population was female (54.9%) with a mean age of 59.1 years (range 18–90, SD 15.6 years) (Table 2). A majority of patients were between the ages of 50 and 79 years (68.6%) and had ASA 2 or 3 status (74.4%). Just under half of procedures were performed in the Northeast (44.0%), followed by the South (27.3%), Midwest (24.0%), with a minority of cases in the West (3.6%). Roughly half of all procedures were colonoscopies (51.4%). Over a third (37.4%) were EGDs, and a minority of cases were ERCPs (6.5%). Regarding anesthesia characteristics, over half of cases were performed with monitored anesthesia care (MAC) or sedation (53.8%) as compared to general anesthesia (36.7%) (Table 2). Other procedural, facility, and anesthesia characteristics are delineated in Table 2.
Table 2.
Patient, Facility, Procedure, and Anesthesia Characteristics of Endoscopic Procedures Performed in the U.S. from 2010 to 2015 Reported in NACOR Database (N= 428,947)
| PATIENT CHARACTERISTICS | ||
|---|---|---|
|
| ||
| VARIABLE | FREQUENCY n (%) | |
| Age | 18 – 49 | 96,341 (22.5) |
| 50 – 64 | 170,312 (39.7) | |
| 65 – 79 | 123,883 (28.9) | |
| > 80 | 37,307 (8.7) | |
| Missing | 1,104 (0.3) | |
|
| ||
| Sex | Female | 235,684 (54.9) |
| Male | 185,510 (43.3) | |
| Missing | 7,753 (1.8) | |
|
| ||
| ASA | 1 | 58,179 (13.6) |
| 2 | 170,134 (39.7) | |
| 3 | 148,992 (34.7) | |
| >=4 | 21,744 (5.1) | |
| Missing | 29,898 (6.9) | |
|
| ||
| FACILITY CHARACTERISTICS | ||
|
| ||
| Total Unique Facilities | 255 | |
|
| ||
| US Region | Northeast | 188,991 (44.0) |
| South | 117,148 (27.3) | |
| Midwest | 102,800 (24.0) | |
| West | 15,23 (3.6) | |
| Missing | 4,778 (1.1) | |
|
| ||
| Facility Type | University Hospital | 79,469 (17.4) |
| Large Community Hospital | 38,299 (9.0) | |
| Medium Community Hospital | 135,084 (31.5) | |
| Small Community Hospital | 12,630 (2.9) | |
| Surgical Center | 33,924 (7.9) | |
| Specialty Hospital/ Other Unspecified | 129,613 (30.2) | |
| Missing | 4,778 (1.1) | |
|
| ||
| Facility Case Volume * | Low (0–9) | 146,990 (34.3) |
| Medium (10–27) | 121,588 (28.3) | |
| High (>27) | 132,314 (30.9) | |
| Missing | 28,055 (6.5) | |
|
| ||
| PROCEDURE CHARACTERISTICS | ||
|
| ||
| Procedure Type | Colonoscopy | 220,500 (51.4) |
| EGD | 160,649 (37.4) | |
| ERCP | 27,721 (6.5) | |
| EUS | 12,246 (2.8) | |
| Flexible Sigmoidoscopy | 5,397 (1.3) | |
| Enteroscopy | 2,434 (0.6) | |
|
| ||
| Outpatient vs. Inpatient | Outpatient | 255,962 (59.7) |
| Inpatient | 77,187 (18.0) | |
| Missing | 95,798 (22.3) | |
|
| ||
| Procedure Status | Elective | 418,337 (97.5) |
| Emergency | 10,610 (2.5) | |
|
| ||
| Day of Week | Weekday | 401,116 (93.5) |
| Weekend | 27,831 (6.5) | |
|
| ||
| Shift | Daytime (7:00–17:00) | 384,684 (89.7) |
| After hours (17:01–06:59) | 44,263 (10.3) | |
|
| ||
| Holiday § | Non-Holiday | 416,163 (97.0) |
| U.S. Holiday | 12,784 (3.0) | |
|
| ||
| ANESTHESIA CHARACTERISTICS | ||
|
| ||
| Anesthesia Type | General | 157,506 (36.7) |
| MAC / Sedation | 230,889 (53.8) | |
| Missing | 40,552 (9.5) | |
|
| ||
| Case Duration (Minutes) + | Short (1–26) | 123,972 (28.9) |
| Medium (27–40) | 123,307 (28.8) | |
| Long (>40) | 119,395 (27.8) | |
| Missing | 62,273 (14.5) | |
Facility case volume defined as number of concurrent cases with anesthesia services taking place at the facility at the time of procedure.
Holiday defined as 11 U.S Government defined holiday including New Year’s Day, Martin Luther King Day, Easter, Memorial Day, Independence Day, Labor Day, Columbus Day, Veterans Day, Thanksgiving, Christmas.
Case duration defined as anesthesia start time to end time.
Abbreviations: American Society of Anesthesiologists (ASA); Esophagogastroduodenoscopy (EGD); Endoscopic retrograde cholangiopancreatography (ERCP); Endoscopic ultrasound (EUS); Monitored anesthesia care (MAC)
Prevalence of Anesthesia Complications
From 2010 to 2015, there were a total of 4441 complications (1.09%) with 1349 serious complications (0.34%). Table 3 provides calculated prevalence of complications stratified by procedure type. Overall, the most common type of complication was post-procedural (3334 cases) including extended stay in the recovery unit, post-procedural nausea/vomiting, inadequate pain control, urinary retention, or unplanned admission. The least common complication was drug-related (0.01%) and a total of 36 deaths (0.01%) occurred in the population. The top 5 ICD-9 diagnostic codes associated with serious complications are listed in Supplementary Table 2.
Table 3.
Prevalence of Anesthesia Complications Among Endoscopic Procedures Stratified by Procedure Type (2010–2015)
| COMPLICATION TYPE | FREQUENCY (n) | TOTAL CASES (N) | RISK (%) | POINT PREVALENCE* PER 100,000 PROCEDURES (95% CI) |
||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ALL | EGD | COLON | ERCP | ALL | EGD | COLON | ERCP | ALL | EGD | COLON | ERCP | ALL | EGD | COLON | ERCP | |
| Serious Complication | 1349 | 511 | 393 | 445 | 400892 | 166260 | 207618 | 27014 | 0.34 | 0.31 | 0.19 | 1.65 | 336.5 (318.8 – 354.9) |
307.4 (281.3–335.1) |
189.3 (171.1–208.9) |
1647.3 (1498.9–1806.2) |
| Any Complication | 4441 | 1647 | 1742 | 1052 | 406285 | 167467 | 211689 | 27129 | 1.09 | 0.98 | 0.82 | 3.89 | 1093.1 (1061.3–1125.5) |
983.5 (936.8–1031.9) |
822.9 (784.9–862.3) |
3877.8 (3651.1–4114.3) |
| Post-Procedure | 3334 | 1243 | 1399 | 692 | 364720 | 147782 | 192073 | 24865 | 0.91 | 0.84 | 0.73 | 2.78 | 914.1 (883.5–945.5) |
841.1 (795.2–889.9) |
728.4 (690.8–767.4) |
2783.0 (2582.1–2995.1) |
| Unplanned Admission | 150 | 76 | 37 | 37 | 337794 | 139928 | 173522 | 24344 | 0.04 | 0.05 | 0.02 | 0.15 | 44.4 (37.6 – 52.1) |
54.3 (4.28–6.79) |
21.3 (15.0–29.4) |
1334.7 (939.9,1839.3) |
| Cardiovascular | 937 | 320 | 280 | 337 | 349249 | 148370 | 174867 | 26012 | 0.27 | 0.22 | 0.16 | 1.30 | 268.2 (251.4–286.0) |
215.7 (192.7–240.6) |
160.1 (141.9–180.0) |
1295.6 (1161.7–1440.5) |
| Respiratory | 200 | 88 | 43 | 69 | 351323 | 150169 | 175131 | 26023 | 0.06 | 0.06 | 0.02 | 0.27 | 56.9 (49.3 – 65.4) |
58.6 (47.0–72.2) |
24.6 (17.8–33.1) |
265.2 (206.4–335.5) |
| Injury | 59 | 30 | 16 | 13 | 262316 | 110077 | 131346 | 20893 | 0.02 | 0.03 | 0.01 | 0.06 | 22.5 (17.1 – 29.0) |
27.3 (18.4–38.9) |
12.2 (6.9–19.8) |
62.2 (33.1–106.4) |
| Neurologic | 53 | 20 | 25 | 8 | 244342 | 105525 | 117180 | 21637 | 0.02 | 0.02 | 0.02 | 0.04 | 21.7 (16.3 – 28.4) |
18.9 (11.6–29.3) |
21.3 (13.8–31.5) |
36.9 (15.9–72.8) |
| Death | 36 | 16 | 8 | 12 | 265707 | 120014 | 121499 | 24194 | 0.01 | 0.01 | 0.01 | 0.04 | 13.6 (9.5 – 18.8) |
13.3 (7.6–21.7) |
6.6 (2.8–13.0) |
49.6 (25.6–86.6) |
| Drug Related | 16 | 5 | 8 | 3 | 211448 | 90288 | 101080 | 20080 | 0.01 | 0.01 | 0.01 | 0.01 | 7.6 (4.3 – 12.3) |
5.5 (1.8–12.9) |
7.9 (3.4–15.6) |
14.9 (3.1–43.7) |
Prevalence calculated as # of complications (frequency) divided by total # of procedures with complete complications reported (total cases at risk for complications).
Predictors of Serious Complications: Unadjusted Bivariate Analysis
Several characteristics were significantly associated with serious complications in the unadjusted bivariate analysis: older individuals (65–79 years), males, ASA 3 class, cases performed in the South, at community hospitals, and higher volume centers (Supplemental Table 3). EGDs (including upper EUS and enteroscopy) had more serious complications than colonoscopies and ERCPs (37.9% vs. 33.0% vs. 29.1% respectively, p<0.001). Longer cases were associated with more serious complications, in addition to cases performed under general anesthesia.
Predictors of Serious Complications: Mixed Effects Multivariable Model
Among the 428,947 endoscopic procedures, 400,892 had complete data recorded for serious complication and 256,909 observations were included in the final multivariable model. Population characteristics for the final model are provided in Supplemental Table 4. In our mixed effects analysis, increasing age, ASA class 4/5, EGD, general anesthesia, overnight shift, longer cases were independently associated with serious complications after adjusting for potential confounders including specific facility, practice and provider identification codes (Table 4 and Supplemental Figure 1). Non-significant variables such as sex and emergent status were kept in the final model given their suspected clinical importance (Table 4).
Table 4.
Unadjusted and Adjusted Associations Between Patient, Facility, Procedure, and Anesthesia Characteristics of Endoscopic Procedures and Serious Anesthesia Complications—Mixed Effects Multivariate Model (N=256,909)*
| VARIABLE | UNADJUSTED OR (95% CI) | ADJUSTED OR (95% CI)* | |
|---|---|---|---|
| Age (years) | 18 – 49 | 1.00 (referent) | 1.00 (referent) |
| 50 – 64 | 1.02 (0.87, 1.20) | 1.37 (1.01, 1.88) | |
| 65 – 79 | 1.58 (1.35, 1.84) | 1.61 (1.16, 2.23) | |
| >=80 | 2.12 (1.75, 2.56) | 1.75 (1.67, 2.63) | |
|
| |||
| Sex | Female | 1.00 (referent) | 1.00 (referent) |
| Male | 1.29 (1.16, 1.43) | 1.08 (0.87, 1.34) | |
|
| |||
| ASA | ASA 1/2 | 1.00 (referent) | 1.00 (referent) |
| ASA 3 | 1.65 (1.47, 1.86) | 1.27 (0.96, 1.68) | |
| ASA 4/5 | 4.10 (3.49, 4.81) | 3.17 (2.15, 4.67) | |
|
| |||
| Location | Northeast | 1.00 (referent) | 1.00 (referent) |
| South | 4.78 (4.20, 5.44) | 1.52 (0.34, 6.85) | |
| Midwest/West | 0.66 (0.54, 0.82) | 0.86 (0.27, 2.75) | |
|
| |||
| Procedure | Colonoscopy | 1.00 (referent) | 1.00 (referent) |
| EGD^ | 1.63 (1.43, 1.85) | 1.44 (1.12, 1.85) | |
| ERCP | 8.83 (7.70, 10.12) | 1.43 (0.98, 2.08) | |
|
| |||
| Facility Type | Community Hospital | 1.00 (referent) | 1.00 (referent) |
| University Hospital | 0.23 (0.19, 0.29) | 1.80 (0.61, 5.36) | |
| Surgical Center | 0.17 (0.11, 0.25) | 1.24 (0.42, 3.69) | |
|
| |||
| Anesthesia Type | MAC / Sedation | 1.00 (referent) | 1.00 (referent) |
| General | 3.09 (2.75, 3.48) | 2.95 (2.08, 4.18) | |
|
| |||
| Shift | Daytime | 1.00 (referent) | 1.00 (referent) |
| Overnight | 8.47 (7.61, 9.43) | 1.56 (1.00, 2.44) | |
|
| |||
| Emergent Status | Elective | 1.00 (referent) | 1.00 (referent) |
| Emergent | 0.75 (0.51, 1.10) | 2.05 (0.84, 5.03) | |
|
| |||
| Case Duration+ | Short | 1.00 (referent) | 1.00 (referent) |
| Medium | 0.95 (0.74, 1.20) | 1.06 (0.75, 1.50) | |
| Long | 3.49 (2.89, 4.23) | 2.44 (1.78, 3.33) | |
|
| |||
| Facility Volume§ | Low | 1.00 (referent) | 1.00 (referent) |
| Medium | 1.12 (0.91, 1.37) | 0.91 (0.63, 1.33) | |
| High | 5.66 (4.85, 6.60) | 1.42 (0.85, 2.37) | |
N is for model adjusted for age, sex, ASA status, location, procedure, facility type, anesthesia type, shift status, emergent status, case duration, facility volume, as well as facility, practice and provider IDs (mixed effects accounting for unique recording behaviors amongst facilities, practices, and providers).
EGD procedures include enteroscopy and endoscopic ultrasound.
Case duration defined as anesthesia start time to end time.
Facility case volume defined as number of concurrent cases with anesthesia services taking place at the facility at the time of procedure.
Abbreviations: American Society of Anesthesiologists (ASA); Esophagogastroduodenoscopy (EGD); Endoscopic retrograde cholangiopancreatography (ERCP); Monitored anesthesia care (MAC)
Patient and Facility Characteristics
In our final mixed effects model, older age was associated with higher odds of serious complications with an OR 1.75 (95% CI 1.67, 2.63) for age >80 years. Patients who were ASA class ≥4 had more than three times the odds of serious complications compared to those who were ASA class 1 or 2 (OR 3.17; 95% CI 2.15, 4.67) (Table 4). On unadjusted analyses, procedures performed in a university setting had a lower odds of serious complication compared to community settings, but when adjusting for unique facility, practice, and provider reporting behaviors (using a mixed effects model), this relationship no longer was significant. Similarly, on unadjusted analyses, serious complications occurred more in the South as compared to the Northeast, and were more commonly reported in high vs. low volume centers; however, this relationship also was no longer significant in the adjusted analyses suggesting that the association with serious complications was likely confounded by varying reporting practices in these settings. Male sex was an independent risk factor for serious complications on unadjusted analyses, but was not significant in the mixed effects model.
Procedural and Anesthesia Characteristics
EGDs had higher odds of serious complication as compared to colonoscopy adjusting for the variables in the mixed effects model (OR 1.44; 95% CI 1.12, 1.85). ERCPs were also associated with higher odds of serious complication in the unadjusted analysis; however, this relationship no longer held true in the multivariable mixed effects model. General anesthesia had about a 3-fold higher odds of serious complications compared to MAC/sedation (OR 2.66; 95% CI 2.12, 3.32). Longer cases (OR 2.44; 95% CI 1.78, 3.33) and cases performed overnight (OR 1.56; 95% CI 1.00, 2.44) also had higher odds of serious complications (Table 4).
Serious Complications Stratified by Procedure Type
Table 5 presents odds ratios for serious complication stratified by procedure type using the same mixed effects model adjusting for age, sex, ASA class, location, facility type, anesthesia type, shift, emergent status, case duration, facility volume, as well as unique facility, practice, and provider IDs. When examining each procedure type separately, older age (65–79 years) was associated with higher odds of serious complications among colonoscopies (OR 2.74; 95% CI 1.65, 4.54), as well as in ERCPs, although this was not significant. ASA class 4/5 was associated with higher odds of serious complications for all procedure types including EGDs (OR 1.96; 95% CI 1.31, 2.92), colonoscopies (OR 2.41; 95% CI 1.05, 5.52), and ERCPs (OR 3.75, 95% CI 1.52, 9.24). ASA status had a stronger association with serious complications among ERCPs with just under 4 times the odds of serious complication as compared to lower ASA categories.
Table 5.
Unadjusted and Adjusted Associations Between Variables and Serious Anesthesia Complications Stratified by Procedure Type (EGD vs. Colonoscopy vs. ERCP)
| VARIABLES | EGD* (n = 111,342)** | COLONOSCOPY (n=153,388) | ERCP (n = 11,827) | ||||
|---|---|---|---|---|---|---|---|
|
| |||||||
| Unadjusted OR (95% CI) | Adjusted OR** (95% CI) | Unadjusted OR (95% CI) | Adjusted OR** (95% CI) | Unadjusted OR (95% CI) | Adjusted OR** (95% CI) | ||
| Age (years) | 18 – 49 | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) |
| 50 – 64 | 1.10 (0.87, 1.40) | 1.17 (0.77, 1.77) | 1.15 (0.83, 1.58) | 1.39 (0.88, 2.18) | 1.57 (1.17, 2.12) | 0.74 (0.31, 1.78) | |
| 65 – 79 | 1.37 (1.08, 1.74) | 1.29 (0.84, 1.99) | 1.74 (1.26, 2.40) | 1.71 (1.06, 2.78) | 2.16 (1.63, 2.12) | 2.02 (0.94, 4.34) | |
| >=80 | 1.70 (1.27, 2.26) | 1.44 (0.84, 2.46) | 1.91 (1.23, 2.95) | 1.14 (0.52, 2.49) | 2.18 (1.58, 3.02 | 2.07 (0.86, 5.03) | |
|
| |||||||
| Sex | Female | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) |
| Male | 1.46 (1.22, 1.74) | 1.35 (0.99, 1.83) | 1.08 (0.88, 1.31) | 0.76 (0.56, 1.03) | 1.29 (1.07, 1.56) | 1.37 (0.80, 2.34) | |
|
| |||||||
| ASA | ASA 1/2 | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) |
| ASA 3 | 1.22 (1.01, 1.48) | 0.91 (0.61, 1.36) | 1.47 (1.19, 1.82) | 1.57 (1.09, 2.24) | 1.65 (1.33, 2.05) | 1.80 (0.88, 3.70) | |
| ASA 4/5 | 2.75 (2.14, 3.54) | 2.74 (1.65, 4.54) | 3.27 (2.26, 4.72) | 2.41 (1.05, 5.52) | 4.08 (3.11, 5.36) | 3.75 (1.52, 9.24) | |
|
| |||||||
| Location | Northeast | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) |
| South | 4.29 (3.48, 5.29) | 1.27 (0.22, 7.23) | 6.14 (4.78, 7.90) | 3.21 (0.53, 19.4) | 5.01 (4.02, 6.24) | 5.18 (0.62, 43.1) | |
| Midwest/West | 0.82 (0.62, 1.11) | 0.72 (0.20, 2.56) | 0.48 (0.30, 0.77) | 0.92 (0.23, 3.71) | 0.64 (0.44, 0.92) | 1.12 (0.20, 6.10) | |
|
| |||||||
| Facility | Community | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) |
| University | 0.33 (0.24, 0.44) | 1.51 (0.43, 5.23) | 0.16 (0.89, 0.29) | 1.35 (0.25, 7.37) | 0.24 (0.16, 0.36) | 0.81 (0.20, 3.35) | |
| Surgical | 0.14 (0.06, 0.31) | 0.47 (0.21, 3.59) | 0.42 (0.27, 0.66) | 1.32 (0.22, 8.04) | NE | NE | |
|
| |||||||
| Anesthesia | Sedation | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) |
| General | 2.47 (2.05, 2.96) | 2.31 (1.47, 3.63) | 1.29 (1.05, 1.59) | 1.48 (0.88, 2.51) | 5.68 (3.39, 9.52) | 1.77 (0.71, 4.38) | |
|
| |||||||
| Shift | Daytime | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) |
| Overnight | 5.59 (4.69, 6.67) | 1.87 (1.06, 3.27) | 8.76 (7.18, 10.7) | 3.10 (1.46, 6.58) | 6.82 (5.61, 8.30) | 0.58 (0.16, 2.12) | |
|
| |||||||
| Status | Non-Emergent | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) |
| Emergent | 0.88 (0.53, 1.47) | 2.38 (0.91, 6.23) | 0.35 (0.11, 1.10) | 1.74 (0.25, 12.2) | 0.84 (0.43, 1.63) | NE | |
|
| |||||||
| Case Duration | Short | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) |
| Medium | 0.88 (0.61, 1.27) | 1.05 (0.64, 1.73) | 0.97 (0.68, 1.36) | 0.83 (0.55, 1.25) | 0.46 (0.23, 0.90) | 0.21 (0.27, 1.61) | |
| Long | 3.81 (2.94, 4.94) | 4.11 (2.73, 6.19) | 1.69 (1.21, 2.34) | 1.08 (0.71, 1.64) | NE | NE | |
|
| |||||||
| Facility Volume | Low | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) | 1.00 (referent) |
| Medium | 0.94 (0.69, 1.27) | 0.76 (0.46, 1.27) | 1.18 (0.83, 1.67) | 0.98 (0.48, 1.97) | 1.42 (0.90, 2.25) | 1.93 (0.85, 4.40) | |
| High | 3.61 (2.88, 4.52) | 1.22 (0.60, 2.46) | 5.67 (4.32, 7.44) | 1.43 (0.63, 3.26) | 4.96 (3.46, 7.13) | 5.56 (1.56, 19.8) | |
EGD procedures include enteroscopy and endoscopic ultrasound.
N is total population in adjusted model. Multivariable mixed effects model excluded non-significant variables day of the week and holiday. For EGD and ERCP, model adjusted for the variables listed above, as well as facility, practice and provider ID (hierarchical modeling) to account for varying reporting practices. For colonoscopy, hierarchical modeling could only account for facility and practice ID; including provider ID was not computationally feasible.
Case duration defined as anesthesia start time to end time.
Facility case volume defined as number of concurrent cases with anesthesia services taking place at the facility at the time of procedure.
Abbreviations: American Society of Anesthesiologists (ASA); Esophagogastroduodenoscopy (EGD); Endoscopic retrograde cholangiopancreatography (ERCP); Monitored anesthesia care (MAC); NE: no estimate due to sparse data
Overnight EGDs and colonoscopies were associated with higher odds of serious complications as compared to daytime procedures (OR 1.87; 95% CI 1.06, 3.27 and OR 3.10; 95% CI, 1.46, 6.58, respectively). For EGDs, general anesthesia was associated with more than double the odds of serious complication as compared to MAC/sedation (OR 2.31; 95% CI 1.47, 3.63). Longer EGD cases had over 4 times the odds of serious complication as compared to shorter duration cases (OR 4.11; 95% CI 2,73, 6.19) Moreover, higher volume centers reported over 5 times the odds of serious complication (OR 5.56; 95% CI 1.56, 19.79).
In a sensitivity analysis separating out different types of EGD procedures, EUS and enteroscopy were not associated with a significantly higher odds of serious complications, and thereby did not explain the increased risk of complications among EGDs overall. In adjusted analyses, EUS had 1.23 times the odds of serious complications (95% CI 0.73, 2.09) and enteroscopy had 2.19 times the odds of serious complications (95% CI 1.00, 4.81) as compared to colonoscopies. In this sensitivity analysis, EGD alone had a 1.45 odds of serious complication (95% CI 1.11, 1.88) and ERCPs had a 1.44 times the odds of complication (95% CI 0.99, 2.09) as compared to colonoscopies, which were similar to the original model reported in Table 4.
Discussion
This study examines data from a large registry of patients undergoing endoscopic procedures with anesthesiology services in a diverse setting of inpatient and outpatient practices across the U.S. Overall, anesthetic complications were uncommonly reported (1%) and serious complications were rare (0.3%). These estimates, although lower than what has been reported in the literature, support the conclusion that ADS during GI endoscopic procedures is generally safe.16,17 This is the first study to investigate other important risk factors for complications including patient, facility, procedure, and anesthesia-specific characteristics. Importantly, we found that serious complications were significantly more common with EGDs as compared to colonoscopies when adjusting for confounders. We also identified that older age, ASA 4/5 status, cases performed overnight and with general anesthesia were independently associated with serious complications. These findings suggest that serious complications are likely driven by a combination of factors including procedure type, patient disease severity, and case complexity.
Our results highlight the growing popularity of anesthesia services for endoscopy over the past decade. In addition to confirming the growing use of anesthesia services outside of the western U.S.,2,18,16 this study also confirmed that sicker patients undergoing EGDs are at higher risk for complications as compared to colonoscopies.10 Longer cases, performed overnight, with general anesthesia were independently associated with serious anesthetic complications. These cases likely involved sicker patients with increased case complexity. Our suspicion is that the driving force for higher odds of serious complications among general anesthesia stems from the sicker, more complex patient population receiving this anesthetic type. However, it should be acknowledged that cases initiated using MAC/sedation may be converted to general anesthesia if certain complications arise, and therefore, reverse causation could also explain this association. Lastly, case duration may reflect the degree of case complexity and would expectedly be associated with higher odds of serious complication. However, case duration may also reflect the time needed to manage serious complications, especially given case duration was measured as anesthesia start time to end time. Therefore, reverse causation is possible for this association.
Of particular interest was our finding that EGDs had a greater association with serious complications than colonoscopies. This has been suggested elsewhere but has been controversial in the literature. One study reported that EGDs with ADS had a 1.33 (95% CI 1.18, 1.50) times the odds of serious adverse events as compared to endoscopist directed sedation adjusting for age, gender, ASA, among other variables.10 Moreover, worse outcomes including 30-day mortality, were reported in emergent endoscopic cases as compared to non-emergent cases using anesthesia services.13 One explanation for the increased odds of complications among upper endoscopies is the close proximity of the airway during these procedures. Serious complications may be driven by respiratory adverse events. In fact, there were a total of 200 respiratory complications in this study—88 occurred in EGDs including EUS and enteroscopy (44%), 69 in ERCPs (35%), and 43 among colonoscopies (21%) bolstering our suspicion that respiratory complications are more common among upper endoscopies and potentially driving higher rates of serious complications (Table 3).
In a sensitivity analysis separating out EUS from EGDs, EUS was not associated with higher risk of serious complications, and thereby did not explain the increased odds of complications among EGDs overall. While EGDs are frequently performed with a native airway technique, at some institutions it is routine to perform endotracheal intubation for ERCP, perhaps mitigating some of the respiratory risk that could be seen with ERCPs. Similarly, looking at complications stratified by procedure type, we did not see a higher odds of serious complication among ERCPs performed under general anesthesia (Table 5). These findings support the notion that general anesthesia (presumably with endotracheal intubation) may mitigate the risk of serious complication among ERCPs.
Among ERCPs, higher volume centers reported over 5 times the odds of serious complications, perhaps because of more robust quality improvement reporting mechanisms for reporting serious complications. Alternatively, higher volume centers may be performing more complex ERCP cases that are prone to more serious complications.
Our calculated prevalence of any complication was 1.09% (1093 complications per 100,000 endoscopic procedures) with serious complications accounting for 0.34% of cases (336.5 complications per 100,000 procedures). These estimates are lower than what has been reported in the literature, likely in part due to under-reporting of complications by anesthesia providers. It is important to note for the composite outcome of interest, not all practices or facilities reported all outcomes, thus the rates of complications likely reflect a lower limit of true complication rates. By comparison, an Australian prospective cohort study of just over 2000 endoscopic procedures reported a 23.0% incidence of significant unplanned events during endoscopic procedures with anesthesia assistance. However, this cohort included sicker individuals with more comorbidities.13 Furthermore, the authors’ definition of complications was more inclusive and comprised common occurrences such as oxygen desaturations <90% not responsive to jaw thrust, hypotension defined as systolic BP < 90 mmHg requiring IV fluid bolus or vasopressor and so forth. Most of these complications are transient, common, and easily treated, and as a result may not be commonly documented by anesthesia providers, hence leading to potential over-estimation of complications in this study.
Our estimates of complication incidence were closer, albeit still lower, than that of a large retrospective observational cohort featuring endoscopic data and calculated propensity-adjusted serious adverse event (SAE) risks.10 In this cohort, total SAEs occurred in 0.20% of ADS colonoscopies and 0.39% of ADS endoscopies and were comparable to frequencies of SAEs among endoscopic directed sedation. Our study only accounted for procedures done in the presence of ADS, and therefore cannot be compared directly to endoscopist directed sedation. However, the authors of this study also found a higher rate of complications during EGDs as compared to colonoscopies, similar to what was found in this study. We acknowledge that a majority of our overall complications included post-procedural events, many of which were not serious complications (e.g. nausea/vomiting, pain). For this reason, we used a more circumscribed composite outcome (serious complications) for our primary outcome in our models. Furthermore, it remains difficult to tease apart whether complications were anesthesia or procedure-related, although the NACOR database focuses specifically on complications reported by an anesthesia provider.
This study has several limitations. While NACOR is the largest anesthesia QI database in the US, there is still missing data from institutions that do not submit information. Complications are voluntarily reported by anesthesia providers, and thus may be subject to under-reporting and reporting bias given certain providers may be less inclined to report their complications.19 Under these circumstances, our incidence of complications may underestimate the true frequency of complications. While markers of disease severity, such as ASA status were investigated, specific data such as laboratory values and body mass index (BMI) were not available in this database. This limited our ability to more fully account for disease severity and case complexity. Lastly, how anesthesia providers chart anesthetics may differ at various institutions leading to potential misclassification of general anesthesia vs. MAC/sedation cases. For example, a mainly Propofol based anesthetic with a native airway may be charted as MAC at one institution, but charted as a general anesthetic at another institution. While there may be slight variation in documentation of anesthesia type, we expect most coding practices to be consistent (especially among intubated cases) given established ASA criteria for general anesthesia, moderate sedation, and deep sedation cases. Lastly, we were unable to look at specific anesthetic drugs used for cases and so cannot comment on complications associated with Propofol specifically.
Many of the limitations of this study were mitigated by the volume of endoscopic cases (>400,000) and the numerous risk factors investigated. NACOR includes a diverse array of inpatient and outpatient practices across multiple geographic regions. This study explored novel predictors of complications including facility type, daytime vs. overnight shift, and case complexity as measured by case duration and facility volume, which have never been investigated before. We also accounted for patient disease severity by investigating ASA class, emergent status, and inpatient/outpatient location for endoscopic procedures. We uncovered significant independent risk factors for serious complications, namely age, ASA class 4/5, EGDs, longer endoscopic cases, overnight cases, and general anesthesia. We expect there was little measurement bias of the various exposure variables (e.g. age, sex, procedure type). Additionally, the measurement bias of the outcome in this cross-sectional study is primarily related to imperfect sensitivity, not imperfect specificity. In this situation, any decrement in sensitivity, if non-differential, will not bias the risk ratio or prevalence ratio.
In conclusion, ADS in endoscopic procedures seems to be safe with low incidence of serious anesthesia complications. Endoscopists, anesthesiology providers, and patients should be aware of the heightened risk of complications among older individuals with more comorbidities, undergoing EGDs, especially during overnight shifts. Future studies need to investigate these potential risk factors for complications in ADS as compared to endoscopist directed conscious sedation.
Supplementary Material
What You Need to Know.
Background:
The risks of anesthesia-directed sedation (ADS) in gastrointestinal endoscopic procedures is not clear; we analyzed national data from multiple centers to determine patterns of use of anesthesia services and risk factors for serious complications.
Findings:
In an analysis of data from the National Anesthesia Clinical Outcomes Registry, we found ADS during endoscopy to be safe, with few serious complications. Risk of ADS complications increased with patients of older age, more severe disease, procedure type, and case complexity.
Implications for Patient Care:
ADS during endoscopy appears to be safe but should be carefully considered for patients who are older or have more severe disease.
Grant Support:
Dr. Lieber is supported by a grant from the National Institutes of Health (T32DK007634) and Dr. Crockett is supported by a grant from the National Institutes of Health (KL2-TR001109).
Abbreviations:
- ADS
Anesthesia directed sedation
- ASA
American Society of Anesthesia
- CRNA
Certified registered nurse anesthetist
- EDS
Endoscopist directed sedation
- EGD
Esophagogastroduodenoscopy
- ERCP
Endoscopic retrograde cholangiopancreatography
- ICD
International Classification of Diseases
- MAC
Monitored anesthesia care
- NACOR
National Anesthesia Clinical Outcomes Registry
Footnotes
Disclosures:
The authors have no conflicts of interest to disclose.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Inadomi JM, Gunnarsson CL, Rizzo JA, et al. Projected increased growth rate of anesthesia professional–delivered sedation for colonoscopy and EGD in the United States: 2009 to 2015. Gastrointest Endosc 2010;72:580–586. [DOI] [PubMed] [Google Scholar]
- 2.Liu H, Waxman DA, Main R, et al. Utilization of Anesthesia Services During Outpatient Endoscopies and Colonoscopies and Associated Spending in 2003–2009. JAMA 2012;307:1178. [DOI] [PubMed] [Google Scholar]
- 3.Cohen LB, Wecsler JS, Gaetano JN, et al. Endoscopic sedation in the United States: results from a nationwide survey. Am J Gastroenterol 2006;101:967–74. [DOI] [PubMed] [Google Scholar]
- 4.Khiani VS, Soulos P, Gancayco J, et al. Anesthesiologist Involvement in Screening Colonoscopy: Temporal Trends and Cost Implications in the Medicare Population. Clin Gastroenterol Hepatol 2012;10:58–64.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hirshman S, Mattke S, Liu H. Anesthesia Service Use and the Uptake of Screening Colonoscopies. Med Care 2017;55:623–628. [DOI] [PubMed] [Google Scholar]
- 6.Cooper GS, Kou TD, Rex DK. Complications Following Colonoscopy With Anesthesia Assistance. JAMA Intern Med 2013;173:551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bielawska B, Hookey LC, Sutradhar R, et al. Anesthesia Assistance in Outpatient Colonoscopy and Risk of Aspiration Pneumonia, Bowel Perforation, and Splenic Injury. Gastroenterology 2017;0. [DOI] [PubMed] [Google Scholar]
- 8.Wernli KJ, Brenner AT, Rutter CM, et al. Risks Associated With Anesthesia Services During Colonoscopy. Gastroenterology 2016;150:888–894. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bielawska B, Day AG, Lieberman DA, et al. Risk Factors for Early Colonoscopic Perforation Include Non-Gastroenterologist Endoscopists: A Multivariable Analysis. Clin Gastroenterol Hepatol 2014;12:85–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Vargo JJ, Niklewski PJ, Williams JL, et al. Patient safety during sedation by anesthesia professionals during routine upper endoscopy and colonoscopy: an analysis of 1.38 million procedures. Gastrointest Endosc 2017;85:101–108. [DOI] [PubMed] [Google Scholar]
- 11.Wadhwa V, Issa D, Garg S, et al. Similar Risk of Cardiopulmonary Adverse Events Between Propofol and Traditional Anesthesia for Gastrointestinal Endoscopy: A Systematic Review and Meta-analysis. Clin Gastroenterol Hepatol 2017;15:194–206. [DOI] [PubMed] [Google Scholar]
- 12.Adeyemo A, Bannazadeh M, Riggs T, et al. Does Sedation Type Affect Colonoscopy Perforation Rates? Dis Colon Rectum 2014;57:110–114. [DOI] [PubMed] [Google Scholar]
- 13.Leslie K, Allen ML, Hessian EC, et al. Safety of sedation for gastrointestinal endoscopy in a group of university-affiliated hospitals: a prospective cohort study. Br J Anaesth 2017;118:90–99. [DOI] [PubMed] [Google Scholar]
- 14.Tillquist MN, Gabriel RA, Dutton RP, et al. Incidence and risk factors for early postoperative reintubations. J Clin Anesth 2016;31:80–89. [DOI] [PubMed] [Google Scholar]
- 15.Whitlock EL, Feiner JR, Chen L. Perioperative Mortality, 2010 to 2014. Anesthesiology 2015;123:1312–1321. [DOI] [PubMed] [Google Scholar]
- 16.Dominitz JA, Baldwin L– M, Green P, et al. Regional Variation in Anesthesia Assistance During Outpatient Colonoscopy Is Not Associated With Differences in Polyp Detection or Complication Rates. Gastroenterology 2013;144:298–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Fagà E, De Cento M, Giordanino C, et al. Safety of propofol in cirrhotic patients undergoing colonoscopy and endoscopic retrograde cholangiography. Eur J Gastroenterol Hepatol 2012;24:70–76. [DOI] [PubMed] [Google Scholar]
- 18.Adams MA, Saleh A, Rubenstein JH. A Systematic Review of Factors Associated With Utilization of Monitored Anesthesia Care for Gastrointestinal Endoscopy. Gastroenterol Hepatol (NY) 2016;12:361–70. [PMC free article] [PubMed] [Google Scholar]
- 19.Dutton RP, Dukatz A. Quality improvement using automated data sources: the anesthesia quality institute. Anesthesiol Clin 2011;29:439–54. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.