Open-access gastroscopy is associated with improved outcomes in gastric cancer

Implementation effect of institutional policy of EGD observation time on neoplasm detection

BACKGROUND AND AIMS

The observation time in EGD is associated with detection rate of premalignant or neoplastic lesions in the upper GI (UGI) tract. The aim of this study was to evaluate an institutional policy of EGD observation time on the detection rate of UGI neoplasms.

METHODS

From July 2017 to March 2019, all endoscopists were requested to comply with our institutional policy of spending more than 3 minutes of observation time in every screening EGD. Observation time was defined as the time from when the endoscope reached the duodenum to when it was withdrawn. We obtained a neoplasm detection rate (NDR) during this period and compared it with that of a baseline period from 2009 to 2015.

RESULTS

During the study period, 30,506 EGDs were performed. Mean subject age was 49.9 ± 10.5 years, and 56.5% were men. All endoscopists achieved an average EGD observation time of more than 3 minutes during this period. Mean observation time was 3:35 ± 0:50, which was significantly longer than the baseline (2:38 ± 0:21, P < .001). NDR was .33%, which was higher than the baseline (.23%, P < .001). Even after adjusting for subjects' age and gender, smoking history, and endoscopists' biopsy sampling rate, prolonged EGD observation time of more than 3 minutes increased the NDR of UGI neoplasms (odds ratio, 1.51; 95% confidence interval, 1.21-1.75).

CONCLUSIONS

This study provides evidence that implementing a protocol of a prolonged observation time could increase NDR. Observation time should be an important quality indicator of the EGD examination.

Longer Observation Time Increases Proportion of Neoplasms Detected by Esophagogastroduodenoscopy

BACKGROUND & AIMS

Esophagogastroduodenoscopy (EGD) is commonly used to detect upper gastrointestinal (GI) neoplasms. However, there is little evidence that longer examination time increases rate of detection of upper GI neoplasia. We investigated the association between length of time spent performing a normal screening EGD and rate of neoplasm detection.

METHODS

We performed a retrospective analysis of data from 111,962 subjects who underwent EGD as part of a comprehensive health-screening program from January 2009 to December 2015 in Korea. Endoscopy findings were extracted from reports prepared by 14 board-certified endoscopists. Endoscopists were classified as fast or slow based on their mean examination time for a normal EGD without biopsy during their first year of the study. All endoscopists used the same endoscopy unit. We obtained findings from histologic analyses of GI biopsies from patient records; positive findings were defined as the detection of neoplasms (esophageal, gastric, or duodenal lesions). We examined the association between examination time and proportions of neoplasms detected. The primary outcome measure was the rate of neoplasm detection for each endoscopist (total number of neoplastic lesions detected divided by the number of subjects screened) and as the proportion of subjects with at least 1 neoplastic lesion.

RESULTS

The mean examination time was 2 minutes 53 seconds. Using 3 minutes as a cutoff, we classified 8 endoscopists as fast (mean duration, 2:38 ± 0:21 minutes) and 6 endoscopists as slow (mean duration, 3:25 ± 0:19 minutes). Each endoscopist's mean examination time correlated with their rate of neoplasm detection (R² = 0.54; P = .046). Fast endoscopists identified neoplasms in the upper GI tract in 0.20% of patients, whereas slow endoscopists identified these in 0.28% of patients (P = .0054). The frequency of endoscopic biopsy varied among endoscopists (range, 6.9%-27.8%) and correlated with rate of neoplasm detection (R² = 0.76; P = .0015). On multivariable analysis, slow endoscopists were more likely to detect gastric adenomas or carcinomas than fast endoscopists (odds ratio, 1.52; 95% CI, 1.17-1.97).

CONCLUSIONS

In a retrospective analysis of data from more than 100,000 subjects who underwent EGD in a screening program, we found slow endoscopists detected a higher proportion of neoplasms than fast endoscopists. Examination time is therefore a useful indicator of quality for EGD.

Poor Outcomes of Gastric Cancer Surgery After Admission Through the Emergency Department

BACKGROUND

Outcomes after nonelective surgery for gastric cancer (GC) are poorly defined. Our objective was to compare outcomes of patients undergoing nonelective GC surgery after admission through the emergency department (EDSx) with patients receiving elective surgery or surgery after planned admission (non-EDSx) nationally.

METHODS

The Nationwide Inpatient Sample (NIS) database was used to examine patients undergoing GC surgery between 2008 and 2012. Demographics and outcomes were compared between EDSx and non-EDSx. Multivariable logistic regression was used to examine predictors of discharge to home.

RESULTS

Of 9279 patients, 1143 (12%) underwent EDSx. They were more likely to be female (42 vs. 35%), nonwhite (56 vs. 33%), aged ≥75 years (40 vs. 26%), in the lowest quartile for household income (31 vs. 25%), have one or more comorbidities (87 vs. 70%), treated at a nonteaching hospital (46 vs. 25%), and have a concomitant diagnosis of obstruction, perforation, or bleeding (30 vs. 6%). They had longer total length of stay (LOS; 16 vs. 9 days), longer median postoperative stays (10 vs. 9 days), higher in-hospital mortality (8 vs. 3%), and were less likely to be discharged home (63 vs. 82%). EDSx was more expensive ($125,300 vs. $83,604). EDSx was associated with a lower likelihood of discharge to home (odds ratio 0.52, 95% CI 0.43-0.62).

CONCLUSIONS

Nationally, 12% of GC surgeries are performed after emergency department admission, which occurs more frequently in vulnerable populations and results in worse outcomes. Understanding factors leading to increased EDSx and developing strategies to decrease EDSx may improve GC surgery outcomes.

Longer examination time improves detection of gastric cancer during diagnostic upper gastrointestinal endoscopy

BACKGROUND & AIMS

It is not clear how the duration of upper endoscopy affects the detection of cancer or premalignant lesions that increase the risk for gastric cancer. We investigated whether the length of time spent performing esophagogastroduodenoscopy (EGD) affects the detection of important pathologic features of the stomach.

METHODS

We collected data from 837 symptomatic patients, during a 3-month period in 2010, who underwent a first diagnostic EGD at a tertiary university hospital in Singapore. Endoscopists were classified as fast or slow based on the mean amount of time it took them to perform a normal EGD examination. We used logistic regression to compare between groups the numbers of intestinal metaplasias, gastric atrophies, dysplasias, and cancers detected, using histologic analysis of biopsy samples collected during endoscopy as the standard.

RESULTS

Of 224 normal endoscopies, the mean duration was 6.6 minutes (range, 2-32 min). When we used 7 minutes as the cut-off time, 8 endoscopists were considered to have short mean examination times (mean duration, 5.5 ± 2.1 min; referred to as fast endoscopists), and 8 endoscopists were considered to have long mean examination times (mean duration, 8.6 ± 4.2 min; referred to as slow endoscopists). Eleven cancers and 81 lesions considered to pose risks for cancer were detected in 86 patients; 1.3% were determined to be cancer, 1.0% were determined to be dysplasia, and 8.7% were determined to be intestinal metaplasia and/or gastric atrophy. Slow endoscopists were twice as likely to detect high-risk lesions as fast endoscopists (odds ratio, 2.50; 95% confidence interval, 1.52-4.12), regardless of whether they were endoscopy staff or trainees. The slow endoscopists also detected 3-fold more neoplastic lesions (cancer or dysplasia; odds ratio, 3.42; 95% confidence interval, 1.25-10.38).

CONCLUSIONS

Endoscopists with mean EGD examination times longer than 7 minutes identified a greater number of high-risk gastric lesions than faster endoscopists. Examination time may be a useful indicator of quality assessment for upper endoscopy. Studies are required to test these findings in different populations.

Performance of American Society for Gastrointestinal Endoscopy guidelines for dyspepsia in Saudi population: Prospective observational study

AIM: To evaluate adherence of primary care physicians (PCPs) to international guidelines when referring patients for upper-gastrointestinal endoscopy (UGE), evaluate the importance of alarm symptoms and the performance of the American Society for Gastrointestinal Endoscopy (ASGE) guidelines in a Saudi population.

METHODS: A prospective, observational cross-sectional study on dyspeptic patients undergoing UGE who were referred by PCPs over a 4 mo period. Referrals were classified as appropriate or inappropriate according to adherence to ASGE guidelines.

RESULTS: Total of 221 dyspeptic patients was enrolled; 161 patients met our inclusion criteria. Mean age was 40.3 years (SD ± 18.1). Females comprised 70.1%. Alarm symptoms included low hemoglobin level (39%), weight loss (18%), vomiting (16%), loss of appetite (16%), difficulty swallowing (3%), and gastrointestinal bleeding (3%). Abnormal endoscopy findings included gastritis (52%), duodenitis (10%), hiatus hernia (7.8%), features suggestive of celiac disease (6.5%), ulcers (3.9%), malignancy (2.6%) and gastroesophageal reflux disease (GERD: 17%). Among patients who underwent UGE, 63% met ASGE guidelines, and 50% had abnormal endoscopic findings. Endoscopy was not indicated in remaining 37% of patients. Among the latter group, endoscopy was normal in 54% of patients. There was no difference in proportion of abnormal endoscopic findings between two groups (P = 0.639).

CONCLUSION: Dyspeptic patients had a low prevalence of important endoscopic lesions, and none of the alarm symptoms could significantly predict abnormal endoscopic findings.

"Drop in" gastroscopy outpatient clinic--experience after 9 months

BACKGROUND

Logistics handling referrals for gastroscopy may be more time consuming than the examination itself. For the patient, "drop in" gastroscopy may reduce uncertainty, inadequate therapy and time off work.

METHODS

After an 8-9 month run-in period we asked patients, hospital staff and GPs to fill in a questionnaire to evaluate their experience with "drop in" gastroscopy and gastroscopy by appointment, respectively. The diagnostic gain was evaluated.

RESULTS

112 patients had "drop in" gastroscopy and 101 gastroscopy by appointment. The number of "drop in" patients varied between 3 and 12 per day (mean 6.5). Mean time from first GP consultation to gastroscopy was 3.6 weeks in the "drop in" group and 14 weeks in the appointment group. The half-yearly number of outpatient gastroscopies increased from 696 before introducing "drop in" to 1022 after (47% increase) and the proportion of examinations with pathological findings increased from 42% to 58%. Patients and GPs expressed great satisfaction with "drop in". Hospital staff also acclaimed although it caused more unpredictable working days with no additional staff.

CONCLUSIONS

"Drop in" gastroscopy was introduced without increase in staff. The observed increase in gastroscopies was paralleled by a similar increase in pathological findings without any apparent disadvantages for other groups of patients. This should legitimise "drop in" outpatient gastroscopies, but it requires meticulous observation of possible unwanted effects when implemented.

Does healthcare infrastructure have an impact on delay in diagnosis and survival?

INTRODUCTION

The objectives of this study were to evaluate whether healthcare infrastructure impacts delay in diagnosis, and to determine whether healthcare infrastructure and delay in diagnosis impacts survival in gastric cancer.

METHODS

Administrative data from 2175 gastric cancer patients was analyzed using two Cox proportional hazard models with (i) delay in diagnosis and (ii) survival as dependent variables. Density of general practitioners, density of gastroenterologists, characteristics of specialty treatment centers, demographic information, and comorbidities were included in the models. Differentiation was made between urban and rural areas.

RESULTS

The likelihood of being diagnosed increased with an increase in general practitioners (p<0.0001) and gastroenterologists (p<0.0001) in rural areas. In urban areas a higher density of general practitioners reduced delay in diagnosis (p=0.0262), while a higher density of gastroenterologists did not (p=0.2480). The number of gastric cancer cases performed in hospital had a positive impact on survival (p<0.0001), while outpatient infrastructure did not.

CONCLUSION

Delay in diagnosis can be reduced by higher availability of general practitioners and gastroenterologists in rural areas. Given the already very high density of physicians in urban areas there is no effect of additional gastroenterologists. As learning effects can be observed with increased hospital volumes, minimum volumes for treatment of gastric cancer may be defined.

The "two-week wait" referral pathway allows prompt treatment but does not improve outcome for patients with oesophago-gastric cancer

INTRODUCTION

The Two Week Wait Referral Service (2WW) has been implemented as a means of fast-tracking patients with suspected upper gastrointestinal cancers for endoscopy. Whether or not it impacts on the outcome of these patients is unclear. The aim of this study was to compare the outcome of patients referred through 2WW with that of patients with oesophago-gastric cancer identified through alternate referral pathways (routine, emergency).

METHODS

The study population was 340 patients with oesophago-gastric carcinoma (gastric 154) diagnosed during the time period 01/2006-12/2007 at University Hospitals of Leicester NHS Trust. Data were collected prospectively by the MDT co-ordinator and analysed retrospectively.

RESULTS

135 of the 340 patients with oesophago-gastric cancer were diagnosed through the 2WW, 115 patients through routine referral pathways, and 90 patients were admitted on an emergency basis. Patients referred through 2WW had a median referral to 1st treatment time of 47 days (routine 79, emergency 28, p < 0.001 all group comparisons). The number of patients treated with potentially curative intent was 37 of 135 for the 2WW, 42 of 115 for the routine referrals and 10 of 90 for patients admitted as emergencies. The corresponding median survivals for the groups were 239 days (2WW), 405 days (routine) and 121 days (emergency), p < 0.001 (log rank).

CONCLUSIONS

Referral by 2WW resulted in more rapid treatment than routine referral but this did not translate into an improvement in survival. This suggests that the targeting of endoscopy to patients with alarm symptoms is flawed and a less selective approach should be promoted if curable cancers are to be detected.

Evaluation and management of dyspepsia

Dyspepsia is a common clinical problem seen by both primary care physicians and gastroenterologists. Initial evaluation should focus on the identification and treatment of potential causes of symptoms such as gastroesophageal reflux disease (GERD), peptic ulcer disease, and medication side effects but also on recognizing those at risk for more serious conditions such as gastric cancer. This manuscript discusses the evaluation and management of dyspepsia including the role of proton-pump inhibitors, treatment of Helicobacter pylori, and endoscopy. Finally, treatment of refractory functional dyspepsia is addressed.

Time-dependent morphologic change in depressed-type early gastric cancer

BACKGROUND

Depressed-type early gastric cancer (EGC) is known to repeat improvement and exacerbation of ulceration during its natural course, forming a "malignant cycle." However, it is difficult to observe the malignant cycle of EGC in clinical practice, and little is known about the clinicopathologic factors associated with this cycle. This study aimed to evaluate the malignant cycle of EGC and to determine the clinicopathologic factors associated with the time-dependent morphologic change of EGC.

METHODS

The medical records of EGC patients treated at the Seoul National University Hospital were retrospectively reviewed with two or more comparable endoscopic photos taken between March 1999 and December 2005. The ulcer stages in EGC were classified by the gastric ulcer stage system and evaluated for time-dependent morphologic changes according to the relevant factors.

RESULTS

In this study, 231 cases of depressed-type EGC were evaluated. At the follow-up endoscopy after a median interval of 23 days, a change in ulcer stage was observed in 66 patients (29%), with improvement in 45 patients (20%) and exacerbation in 21 patients (9%). Of the 177 patients who demonstrated an active or healing stage, 45 (25%) showed improvement and 10 (6%) exhibited exacerbation. The multivariate analysis showed that the usage of antisecretory medications, mucosal cancer, and a longer interval between the two endoscopic examinations were the factors that had a significant association with improvement of the ulcer stage.

CONCLUSIONS

The malignant cycle was frequently observed in depressed-type EGC. The improvement of ulceration in EGC was time dependent and associated with the usage of antisecretory medication and the depth of invasion.

Role of symptoms in diagnosis and outcome of gastric cancer

Gastric cancer is one of the most common cancers and the second most common cause of cancer deaths worldwide. Apart from Japan, where screening programmes have resulted in early diagnosis in asymptomatic patients, in most countries the diagnosis of gastric cancers is invariably made on account on dyspeptic and alarm symptoms, which may also be of prognostic significance when reported by the patient at diagnosis. However, their use as selection criteria for endoscopy seems to be inconsistent since alarm symptoms are not sufficiently sensitive to detect malignancies. In fact, the overall prevalence of these symptoms in dyspeptic patients is high, while the prevalence of gastro-intestinal cancer is very low. Moreover, symptoms of early stage cancer may be indistinguishable from those of benign dyspepsia, while the presence of alarm symptoms may imply an advanced and often inoperable disease. The features of dyspeptic and alarm symptoms may reflect the pathology of the tumour and be of prognostic value in suggesting site, stage and aggressiveness of cancer. Alarm symptoms in gastric cancer are independently related to survival and an increased number, as well as specific alarm symptoms, are closely correlated to the risk of death.Dysphagia, weight loss and a palpable abdominal mass appear to be major independent prognostic factors in gastric cancer, while gastro-intestinal bleeding, vomiting and also duration of symptoms, do not seem to have a relevant prognostic impact on survival in gastric cancer.

Relative value of repeat gastric ulcer surveillance gastroscopy in diagnosing gastric cancer

BACKGROUND

Gastric cancer can present with the endoscopic appearances of a benign gastric ulcer (GU). Opinion remains divided on the need for follow-up of patients diagnosed with GU, and the aim of this study was to examine the long-term outcomes of patients whose GU proved malignant on follow-up gastroscopy.

METHODS

Between October 1, 1995, and September 30, 2003, 25,579 gastroscopies were performed in one unit. These identified 544 patients with apparently benign GU, of whom 277 (51%) underwent 334 elective follow-up endoscopies. Twelve of these patients (4.3%) were shown to have a malignant ulcer; their outcomes were compared to those of the 296 other patients diagnosed with gastric cancers in this time frame.

RESULTS

The patients in the GU cancer group had earlier stage disease (stage I, 33% vs 6.4%; chi2 = 11.2; DF1; P = 0.001), and were more likely to undergo R0 gastrectomy (50% vs 30%; chi2 = 2.064; DF1; P = 0.151) and to survive long term (46% vs 16%; log-rank chi2, 5.79; DF1; P = 0.0162) than patients in the comparison cohort.

CONCLUSION

Gastroscopic follow-up of 50 patients with an apparently benign GU will identify 1 patient with a malignancy destined to survive for 5 years following R0 gastrectomy. This justifies the diagnostic effort of repeat gastroscopy to ensure complete healing of GU.

Use of alarm symptoms to select dyspeptics for endoscopy causes patients with curable esophagogastric cancer to be overlooked

BACKGROUND

In August 2004, the United Kingdom Department of Health advisory body published dyspepsia referral guidelines for primary care practitioners. These guidelines advised empiric treatment with antisecretory medications and referral for endoscopy only in the presence of alarm symptoms. The current study aimed to evaluate the effect of these guidelines on the detection of esophagogastric cancer.

METHODS

The study reviewed a prospectively compiled database of 4,018 subjects who underwent open access gastroscopy during the years 1990 to 1998. The main outcome measures for the study were cancer detection rates, International Union Against Cancer (UICC) stage, and survival.

RESULTS

Gastroscopy identified esophagogastric carcinoma in 123 (3%) of the 4,018 subjects. Of these 123 patients, 104 (85%) with esophagogastric cancer had "alarm" symptoms (anemia, mass, dysphagia, weight loss, vomiting) and would have satisfied the referral criteria. The remaining 15% would not have been referred for initial endoscopic assessment because their symptoms were those of uncomplicated "benign" dyspepsia. The patients with "alarm" symptoms had a significantly more advanced tumor stage (metastatic disease in 47% vs 11%; p < 0.001), were less likely to undergo surgical resection (50% vs 95%; p < 0.001), and had a poorer survival (median, 11 vs 39 months; p = 0.01) than their counterparts without such symptoms.

CONCLUSIONS

The use of alarm symptoms to select dyspeptics for endoscopy identifies patients with advanced and usually incurable esophagogastric cancer. Patients with early curable cancers often have only dyspeptic symptoms, and their diagnosis will be delayed until the symptoms of advanced cancer develop.

Prognostic significance of alarm symptoms in patients with gastric cancer

BACKGROUND

The aim of this study was to determine the incidence and spectrum of alarm symptoms in patients with newly diagnosed gastric cancer, and to examine the relationship between symptoms and outcome.

METHODS

Three hundred consecutive patients with gastric adenocarcinoma were studied prospectively. The outcomes of 40 patients (13.3 per cent) without alarm symptoms (21 men; median age 69 years) were compared with those of the 260 patients (86.7 per cent) with alarm symptoms (175 men; median age 72 years).

RESULTS

It was possible to perform an R0 gastrectomy more often in patients without alarm symptoms (21 patients; 52 per cent) than in those with alarm symptoms (71 patients; 27.3 per cent) (chi(2) = 10.35, 1 d.f., P = 0.001). The cumulative survival rate at 5 years was 38 per cent for patients without alarm symptoms versus 15.0 per cent for those with alarm symptoms (chi(2) = 10.18, 1 d.f., P = 0.001). In a multivariate analysis, distant metastasis (hazard ratio (HR) 2.73 (95 per cent confidence interval (c.i.) 2.04 to 3.66); P < 0.001), overall stage of cancer (HR 1.83 (95 per cent c.i. 1.53 to 2.19); P < 0.001) and persistent vomiting at diagnosis (HR 1.66 (95 per cent c.i. 1.26 to 2.18); P < 0.001) were independently associated with length of survival.

CONCLUSION

Alarm symptoms are absent in a significant minority of patients with gastric cancer at diagnosis; these patients stand a better chance of curative surgery and long-term survival than those with alarm symptoms.