Retained gastric food contents are frequently encountered during upper endoscopy. Depending on the patient population and clinical setting, retained gastric contents have been reported in 0.3%–19.0% of upper endoscopies (1–3). Given the rising incidence and morbidity of gastroparesis, the incidence of retained gastric content encountered during upper endoscopy may also increase (4). Retained gastric food content may lead to suboptimal examination, affecting accurate diagnosis and delaying therapeutic interventions. This in turn may lead to a need for repeat procedures, which is associated with additional procedural risk and cost. Finally, aspiration of gastric food content during the period of sedation may lead to potentially life-threatening respiratory adverse events including aspiration pneumonia and chemical pneumonitis (5–7).
While previous studies have primarily focused on identifying risk factors for retained gastric food content (1), the effect of retained gastric contents during endoscopy on diagnostic accuracy, outcomes of therapeutic interventions, and patient safety remains unknown. We hypothesize that retained gastric content encountered during endoscopy may reduce diagnostic accuracy and increase the risk of adverse respiratory events related to aspiration. Furthermore, given the lack of specific recommendations from society guidelines, substantial variability among individual endoscopist practice patterns is expected. The primary aims of this study were to evaluate endoscopists' practice patterns by assessing the proportion of patients who were able to undergo a complete endoscopic examination to the second portion of the duodenum despite retained food contents and the incidence of respiratory adverse events after endoscopy. Predictors associated with complete examination and respiratory adverse events were also evaluated in this patient population.
METHODS Study designLoma Linda University Institutional Review Board approval was obtained before initiating this retrospective study. Consecutive patients older than 18 years with retained gastric food content observed during first-time upper endoscopy at Loma Linda University Medical Center from January 2016 to March 2021 were included in the study.
Upper endoscopy was performed after >6 hours of fasting of solids and >2 hours from clear liquid based on institutional gastrointestinal (GI) laboratory and anesthesia protocol. Search terms (“food,” “retention,” “retained,” “bezoar,” and “debris”) were used to identify the patient population meeting study criteria from an internal endoscopy database. Procedure reports were subsequently reviewed by 2 independent researchers to verify the presence of retained gastric contents encountered during endoscopy. Exclusion criteria included patients who were pregnant, had received prior foregut surgery, or were already on mechanical ventilation before planned endoscopy. In patients who had multiple procedures, the first procedure at the institution during the study period was considered the index procedure. Medical records including radiologic studies and endoscopic reports were carefully reviewed to characterize the patient's clinical course after endoscopy. Data including demographic information, endoscopist performing the procedure, American Society of Anesthesiologists classification, inpatient vs outpatient status, location of procedure, indication for procedure, type of sedation or anesthesia, and therapeutic interventions performed were collected. Large gastric food burden was defined as large amounts of gastric contents unable to be cleared after endoscopic lavage, which was confirmed by an independent review of endoscopic photographs. Small gastric food burden was defined as gastric content mostly cleared after endoscopic lavage. Retained esophageal food content was defined as the presence of concomitant esophageal food content in addition to gastric contents. Blood content alone in patients with upper GI bleeding was not considered retained gastric food content.
Study endpointsPrimary endpoints included complete examination defined as successful endoscopic advancement to the second portion of the duodenum and respiratory adverse events within 30 days after endoscopy. Respiratory adverse events included any episode of acute respiratory symptoms associated with abnormal radiographic findings after endoscopy. Aspiration pneumonia was defined as the development of respiratory adverse event necessitating antimicrobial therapy. Secondary outcomes included incidence of repeat endoscopy, 30-day hospitalization among those who received outpatient endoscopy, and 30-day mortality after index endoscopy. Repeat endoscopy was defined by the need for repeat procedure within 30 days of the index procedure because of suboptimal examination or inability to perform intended therapy.
Statistical analysisContinuous variables were reported as mean ± SD, and categorical variables were reported as frequency (%). Individual endoscopist's complete examination rates were calculated and then categorized into quintiles in relation to other endoscopists. Univariate logistic regression analyses were used to determine associations between baseline demographic factors and primary outcomes. Significant covariates with P value <0.1 in univariate analysis were included in multivariate logistic regression analyses to derive adjusted odds ratios (aORs) and confidence intervals (CIs). All significance tests were 2-sided, and the significance level was set at 0.05. Statistical analyses were performed using SAS software, version 9.4 (SAS Institute, Cary, NC).
RESULTS Study populationDuring the study period, 730 (4.1%) of 17,868 patients who received first-time upper endoscopy at Loma Linda University Medical Center were found to have retained gastric food content. After excluding patients who had altered surgical anatomy or who were already mechanically ventilated before the procedure, 629 (3.5%) met inclusion criteria (Figure 1). The mean age of patients was 55 ± 17 years, 310 (49.2%) were male, and 283 (44.9%) received endoscopy during hospitalization. Furthermore, 30 (4.8%) patients had a previous diagnosis of gastroparesis before outpatient (21, 3.3%) or inpatient (9, 1.4%) endoscopy; 9 received prokinetics before the procedure. A median of 20 (range 1–77) procedures demonstrating retained gastric food content were performed by 26 endoscopists. The most common primary indications for endoscopy included dysphagia in 143 (22.7%), upper GI bleeding in 133 (21.1%), and vomiting in 75 (12.0%) (Table 1). Procedural characteristics are summarized in Table 2. Moderate sedation, anesthesiologist-assisted sedation, and general anesthesia were performed in 506 (80.5%), 16 (2.5%), and 107 (17.0%) patients, respectively. The median duration of procedure was 22 minutes (range 2–180). Large and small gastric food burden was observed in 256 (40.7%) and 373 (59.3%) patients, respectively. Seventy-five (11.9%) patients were found to have some degree of retained food in the esophagus including 10 with achalasia. One hundred sixty-three (25.9%) patients received planned therapeutic interventions including esophageal dilation in 79 (12.6%), nonvariceal hemostasis in 22 (3.5%), and variceal band ligation in 13 (2.1%). Furthermore, 238 (37.8%) patients received mucosal biopsies including biopsies of the esophagus in 88 (14.0%), stomach in 175 (27.8%), and duodenum in 91 (14.5%) (Table 2).
Figure 1.:Flowchart of patient selection. EGD, esophagogastroduodenoscopy.
Table 1. - Baseline characteristics of patients with gastric food content encountered during endoscopy N = 629 Age, yr, mean ± SD 54.8 ± 16.5 Male, n (%) 310 (49.2) Race/ethnicity, n (%) White 469 (74.6) African American 65 (10.3) Latino 43 (6.8) Asian 33 (5.2) Others 19 (3.0) ASA class, n (%) 1 48 (7.6) 2 270 (42.9) 3 288 (45.8) 4 23 (3.7) History of gastroparesis, n (%) 30 (4.8) Primary indications for EGD, n (%) Dysphagia 143 (22.7) Upper GI bleeding 133 (21.1) Vomiting 75 (12.0) Dyspepsia 58 (9.2) Screen for varices 64 (10.2) GERD symptoms 42 (6.7) Anemia 29 (4.6) Diarrhea 16 (2.5) History of gastric ulcer 15 (2.4) Weight loss 11 (1.7) Others 43 (6.8) Hospitalization, n (%) 283 (44.9)ASA, American Society of Anesthesiologist; EGD, esophagogastroduodenoscopy; GERD, gastroesophageal reflux disease; GI, gastrointestinal.
Of the 629 patients with retained gastric contents encountered during endoscopy, 534 (84.9%) received complete examination to the second portion of duodenum (Table 3). Fifty-eight (9.2%) patients required repeat endoscopy within 30 days of index procedure because of suboptimal examination or postponed intervention. Forty-two (6.7%) patients developed adverse events after endoscopy including 41 (6.5%) who received inpatient and 1 (0.2%) who received outpatient procedures. Of the 42 patients, 41 (6.5%) with retained gastric contents experienced a respiratory adverse event within 30 days, while 1 (0.2%) patient developed an esophageal microperforation after esophageal stricture dilation and was treated conservatively with antibiotics. Respiratory adverse events included aspiration pneumonia requiring antibiotics in 30 (4.8%) patients, respiratory events not requiring antibiotics in 10 (1.6%) patients, and airway obstruction due to obstructive mucous plugging in 1 (0.2%) patient. Of the 41 patients who developed respiratory adverse events, 21 (51.2%) developed respiratory failure requiring intensive care for ventilatory support and 10 (23.8%) died during the hospitalization course. Only 1 patient who received outpatient procedure developed respiratory adverse event. In that case, a 75-year-old woman with nonalcoholic steatohepatitis cirrhosis who was undergoing screening endoscopy for esophageal varices was found to have large amounts of food contents in her stomach. She was hospitalized 2 days later for tense ascites and dyspnea with imaging consistent with chemical pneumonitis and later recovered with diuretics and oxygen therapy.
Table 3. - Clinical outcomes N = 629 Complete examination to the second portion of duodenum, n (%) 534 (84.9) Repeat EGD, n (%) 58 (9.2) Mean days to repeat EGD ± SD 8.8 ± 8.3 Total adverse events related to endoscopy, n (%) 42 (6.7) Respiratory adverse events 41 (6.5) Aspiration pneumonia 30 (4.8) Airway obstruction 1 (0.2) Esophageal perforation 1 (0.2) 30-d hospitalization, n (%)a 18/346 (5.2) 30-d mortality, n (%) 15 (2.4)EGD, esophagogastroduodenoscopy.
aRate of 30-day hospitalization seen in outpatient cases.
Of the 346 patients who received outpatient endoscopy, 18 (5.2%) required hospitalization within 30 days including 1 for respiratory adverse events. Overall, 15 (2.4%) patients died within 30 days of the procedure.
Predictors of complete examinationThe median rate of complete examination among 26 endoscopists was 85.1% (interquartile range, 77%–97%). On univariate analysis, the degree of gastric food contents (74.6% vs 92.2%, P < 0.0001), duration of procedure ≥20 minutes (87.0% vs 82.0%, P = 0.09), obtaining mucosal biopsies (93.7% vs 79.5%, P < 0.0001), and complete examination rate per endoscopists by quintile (70.3%–98.0%, P < 0.0001) were associated with complete examination to the second portion of the duodenum (Table 4). On multivariate analysis, large compared with small gastric food burden was associated with a decreased likelihood of complete examination (aOR 0.26, 95% CI 0.16–0.43), while mucosal biopsies obtained during endoscopy was associated with an increased likelihood of complete examination (aOR 2.54, 95% CI 1.38–4.69). Furthermore, when analyzing individual endoscopist complete examination rates by quintiles, patients who received procedures by endoscopists in the first (aOR 15.64, 95% CI 4.58–53.39), second (aOR 3.81, 95% CI 1.64–8.87), and third quintiles (aOR 2.29, 95% CI 1.18–4.47) were more likely to receive a complete examination compared with those who received procedures in the fifth quintile.
Table 4. - Predictors of deep duodenal intubation in patients with gastric food content encountered during endoscopy Univariate Multivariatea N = 629aOR, adjusted odds ratio; ASA, American Society of Anesthesiologists; CI, confidence interval; GI, gastrointestinal.
aMultivariate logistic regression analysis was conducted when significant covariates were seen in univariate model.
bRate of examination completeness was stratified into quintiles based on how often endoscopists achieved deep duodenum intubation.
cSignifies statistical significance.
On univariate analysis, the American Society of Anesthesiologists class (range 0%–17.4%, P = 0.002), inpatient status (14.1% vs 0.3%, P < 0.0001), upper GI bleeding (16.7% vs 3.8%, P < 0.0001), degree of gastric food content (10.2% vs 4.0%, P = 0.002), and obtaining mucosal biopsies during procedure (3.8% vs 8.2%, P = 0.03) were associated with increased respiratory adverse events (Table 5). Conversely, the type of anesthesia, endoscopic interventions other than biopsies, concurrent esophageal food content, and the duration of the procedure were not associated with odds of respiratory adverse events. On multivariate analysis, inpatient status during endoscopy (aOR 37.78, 95% CI 4.94–288.99) and large gastric food content (aOR 2.14, 95% CI 1.09–4.20) were associated with an increased likelihood of respiratory adverse events.
Table 5. - Pulmonary adverse events in patients with gastric food content encountered during endoscopy Univariate Multivariatea N = 629aOR, adjusted odds ratio; ASA, American Society of Anesthesiologists; CI, confidence interval; GI, gastrointestinal.
aMultivariate logistic regression analysis was conducted when significant covariates were seen in univariate model.
bSignifies statistical significance.
In this single-center US study of patients undergoing upper endoscopy, unexpected gastric food content occurred in approximately 1 in 30 procedures after excluding those with anatomic foregut alterations or prior airway protection. Although deep duodenal intubation was achieved in most (84.8%) patients, substantial variation in complete examination rates was observed among individual endoscopists. The presence of large food burden was associated with decreased odds while obtaining mucosal biopsies during procedure was associated with increased odds of complete examination after adjusting for endoscopist-specific completion rate. Finally, respiratory adverse events occurred in 7% of the study population but
留言 (0)