Association of Intensive Endoscopic Screening Burden With Gastric Cancer Detection | Cancer Screening, Prevention, Control | JAMA Network Open | JAMA Network
[Skip to Content]
[Skip to Content Landing]
Figure 1.  Flow Diagram for Selection of the Screening-Eligible Population
Flow Diagram for Selection of the Screening-Eligible Population

Overall, 21 535 222 participants underwent upper endoscopy for gastric cancer screening. A total of 76 666 (0.77%) and 78 669 (0.68%) participants for each cycle screened positive. Of participants who had negative results based on the upper endoscopy, 7276 (0.07%) and 4531 (0.04%) for each respective cycle were diagnosed with gastric cancer within 1 year of endoscopic examination. UGI indicates upper gastrointestinal.

Figure 2.  Ratio of Monthly Screenings and Monthly Detection Rates for Gastric Cancer in the Korean National Cancer Screening Program
Ratio of Monthly Screenings and Monthly Detection Rates for Gastric Cancer in the Korean National Cancer Screening Program

A, Ratio of the monthly screenings based on age grouping for the 2013-2014 cycle and the 2015-2016 cycle. B, Detection rates per 100 screenings decreased toward the end of the year, especially in December. The error bars indicate 95% CIs.

Table 1.  Characteristics of Screening-Eligible Population Total, the 2013-2014 and 2015-2016 Korean National Cancer Screening Program Cycles
Characteristics of Screening-Eligible Population Total, the 2013-2014 and 2015-2016 Korean National Cancer Screening Program Cycles
Table 2.  Estimation of Screening Month Effect Size Based on the Negative Binomial Regressiona
Estimation of Screening Month Effect Size Based on the Negative Binomial Regressiona
Table 3.  Logistic Regression Analysis of Risk Factors for Gastric Cancer Detection
Logistic Regression Analysis of Risk Factors for Gastric Cancer Detection
1.
Ferlay  J, Soerjomataram  I, Dikshit  R,  et al.  Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012.   Int J Cancer. 2015;136(5):E359-E386. doi:10.1002/ijc.29210 PubMedGoogle ScholarCrossref
2.
Bray  F, Ferlay  J, Soerjomataram  I, Siegel  RL, Torre  LA, Jemal  A.  Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries.   CA Cancer J Clin. 2018;68(6):394-424. doi:10.3322/caac.21492 PubMedGoogle ScholarCrossref
3.
Jun  JK, Choi  KS, Lee  HY,  et al.  Effectiveness of the Korean National Cancer Screening Program in reducing gastric cancer mortality.   Gastroenterology. 2017;152(6):1319-1328.e7. doi:10.1053/j.gastro.2017.01.029 PubMedGoogle ScholarCrossref
4.
Tashiro  A, Sano  M, Kinameri  K, Fujita  K, Takeuchi  Y.  Comparing mass screening techniques for gastric cancer in Japan.   World J Gastroenterol. 2006;12(30):4873-4874. doi:10.3748/wjg.v12.i30.4873PubMedGoogle Scholar
5.
Lee  HY, Park  EC, Jun  JK, Choi  KS, Hahm  MI.  Comparing upper gastrointestinal X-ray and endoscopy for gastric cancer diagnosis in Korea.   World J Gastroenterol. 2010;16(2):245-250. doi:10.3748/wjg.v16.i2.245 PubMedGoogle ScholarCrossref
6.
Leung  WK, Wu  MS, Kakugawa  Y,  et al; Asia Pacific Working Group on Gastric Cancer.  Screening for gastric cancer in Asia: current evidence and practice.   Lancet Oncol. 2008;9(3):279-287. doi:10.1016/S1470-2045(08)70072-X PubMedGoogle ScholarCrossref
7.
Zhang  X, Li  M, Chen  S,  et al.  Endoscopic screening in Asian countries is associated with reduced gastric cancer mortality: a meta-analysis and systematic review.   Gastroenterology. 2018;155(2):347-354.e9. doi:10.1053/j.gastro.2018.04.026 PubMedGoogle ScholarCrossref
8.
Shin  HR, Won  YJ, Jung  KW,  et al; Members of the Regional Cancer Registries.  Nationwide cancer incidence in Korea, 1999-2001; first result using the national cancer incidence database.   Cancer Res Treat. 2005;37(6):325-331. doi:10.4143/crt.2005.37.6.325 PubMedGoogle ScholarCrossref
9.
Lee  S, Jun  JK, Suh  M,  et al.  Gastric cancer screening uptake trends in Korea: results for the National Cancer Screening Program from 2002 to 2011: a prospective cross-sectional study.   Medicine (Baltimore). 2015;94(8):e533. doi:10.1097/MD.0000000000000533 PubMedGoogle Scholar
10.
National Health Insurance Service.  2013 National Health Screening Statistical Yearbook. National Health Insurance Service; 2014.
11.
National Health Insurance Service.  2014 National Health Screening Statistical Yearbook. National Health Insurance Service; 2015.
12.
National Health Insurance Service.  2015 National Health Screening Statistical Yearbook. National Health Insurance Service; 2016.
13.
National Health Insurance Service.  2016 National Health Screening Statistical Yearbook. National Health Insurance Service; 2017.
14.
National Health Insurance Service.  2017 National Health Screening Statistical Yearbook. National Health Insurance Service; 2018.
15.
National Health Insurance Service.  2018 National Health Screening Statistical Yearbook. National Health Insurance Service; 2019.
16.
Harewood  GC, Chrysostomou  K, Himy  N, Leong  WL.  Impact of operator fatigue on endoscopy performance: implications for procedure scheduling.   Dig Dis Sci. 2009;54(8):1656-1661. doi:10.1007/s10620-008-0549-7 PubMedGoogle ScholarCrossref
17.
Kaneshiro  M, Ho  A, Chan  M, Cohen  H, Spiegel  BM.  Colonoscopy yields fewer polyps as the day progresses despite using social influence theory to reverse the trend.   Gastrointest Endosc. 2010;72(6):1233-1240. doi:10.1016/j.gie.2010.08.034 PubMedGoogle ScholarCrossref
18.
Chan  MY, Cohen  H, Spiegel  BM.  Fewer polyps detected by colonoscopy as the day progresses at a Veteran’s Administration teaching hospital.   Clin Gastroenterol Hepatol. 2009;7(11):1217-1223. doi:10.1016/j.cgh.2009.07.013 PubMedGoogle ScholarCrossref
19.
Almadi  MA, Sewitch  M, Barkun  AN, Martel  M, Joseph  L.  Adenoma detection rates decline with increasing procedural hours in an endoscopist’s workload.   Can J Gastroenterol Hepatol. 2015;29(6):304-308. doi:10.1155/2015/789038 PubMedGoogle ScholarCrossref
20.
Rex  DK.  Quality in colonoscopy: cecal intubation first, then what?   Am J Gastroenterol. 2006;101(4):732-734. doi:10.1111/j.1572-0241.2006.00483.x PubMedGoogle ScholarCrossref
21.
Lee  A, Jensen  CD, Marks  AR,  et al.  Endoscopist fatigue estimates and colonoscopic adenoma detection in a large community-based setting.   Gastrointest Endosc. 2017;85(3):601-610.e2. doi:10.1016/j.gie.2016.09.033 PubMedGoogle ScholarCrossref
22.
Hamashima  C.  Current issues and future perspectives of gastric cancer screening.   World J Gastroenterol. 2014;20(38):13767-13774. doi:10.3748/wjg.v20.i38.13767 PubMedGoogle ScholarCrossref
23.
Choi  KS, Jun  JK, Park  EC,  et al.  Performance of different gastric cancer screening methods in Korea: a population-based study.   PLoS One. 2012;7(11):e50041. doi:10.1371/journal.pone.0050041 PubMedGoogle Scholar
24.
Karimi  P, Islami  F, Anandasabapathy  S, Freedman  ND, Kamangar  F.  Gastric cancer: descriptive epidemiology, risk factors, screening, and prevention.   Cancer Epidemiol Biomarkers Prev. 2014;23(5):700-713. doi:10.1158/1055-9965.EPI-13-1057 PubMedGoogle ScholarCrossref
25.
Saposnik  G, Baibergenova  A, Bayer  N, Hachinski  V.  Weekends: a dangerous time for having a stroke?   Stroke. 2007;38(4):1211-1215. doi:10.1161/01.STR.0000259622.78616.ea PubMedGoogle ScholarCrossref
26.
Bendavid  E, Kaganova  Y, Needleman  J, Gruenberg  L, Weissman  JS.  Complication rates on weekends and weekdays in US hospitals.   Am J Med. 2007;120(5):422-428. doi:10.1016/j.amjmed.2006.05.067 PubMedGoogle ScholarCrossref
27.
Hendey  GW, Barth  BE, Soliz  T.  Overnight and postcall errors in medication orders.   Acad Emerg Med. 2005;12(7):629-634. doi:10.1197/j.aem.2005.02.009 PubMedGoogle ScholarCrossref
28.
Maung  R.  Pathologists' workload and patient safety.   Diagn Histopathol. 2016;22(8):283-287. doi:10.1016/j.mpdhp.2016.07.004 Google ScholarCrossref
29.
Sanaka  MR, Shah  N, Mullen  KD, Ferguson  DR, Thomas  C, McCullough  AJ.  Afternoon colonoscopies have higher failure rates than morning colonoscopies.   Am J Gastroenterol. 2006;101(12):2726-2730. doi:10.1111/j.1572-0241.2006.00887.x PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Views 712
    Citations 0
    Original Investigation
    Gastroenterology and Hepatology
    January 7, 2021

    Association of Intensive Endoscopic Screening Burden With Gastric Cancer Detection

    Author Affiliations
    • 1Department of Gastroenterology, Ajou University School of Medicine, Suwon, Republic of Korea
    • 2Stomach Cancer Center, Ajou University Hospital, Suwon, Republic of Korea
    • 3Department of Biomedical Informatics, Ajou University School of Medicine, Suwon, Republic of Korea
    • 4Office of Biostatistics, Ajou Research Institute for Innovative Medicine, Ajou University Medical Center, Suwon, Republic of Korea
    • 5Department of Medical Sciences, Biomedical Informatics, Graduate School of Ajou University, Suwon, Republic of Korea
    • 6Department of Occupational and Environmental Medicine, Ajou University School of Medicine, Suwon, Republic of Korea
    JAMA Netw Open. 2021;4(1):e2032542. doi:10.1001/jamanetworkopen.2020.32542
    Key Points

    Question  Is monthly variation in gastric cancer screening rates associated with changes in rates of cancer detection?

    Findings  This cohort study of 21 535 222 participants from the Korean National Cancer Screening Program database found that the monthly number of endoscopic examinations increased toward the end of the year. However, the gastric cancer detection rates were negatively and significantly associated with calendar month.

    Meaning  The findings of this cohort study suggest the importance of the implementation of policy guidelines to strengthen the merits of the screening program, including a reduction in gastric cancer-related mortality.

    Abstract

    Importance  The rapidly increasing number of gastric cancer examinations performed over a short period might influence screening performance. Accessing the association between calendar month and gastric cancer detection rates might improve policy and guide institutional support.

    Objective  To evaluate the association between the increased number of examinations over a certain period and gastric cancer detection rates among a large population included in the Korean National Cancer Screening Program (KNCSP).

    Design, Setting, and Participants  This retrospective, population-based cohort study used data from the KNCSP comprising 26 765 665 men and women aged 40 years or older who participated in the screening program between January 1, 2013, and December 31, 2016. Data were analyzed from November 1, 2019, to March 31, 2020.

    Exposures  Gastric cancer screening with endoscopy.

    Main Outcomes and Measures  The primary outcome was monthly gastric cancer detection rates in the KNCSP. A negative binomial regression model was used to evaluate the association between the screening month and detection rates.

    Results  In total, 21 535 222 individuals underwent endoscopy (mean [SD] age, 55.61 [10.61] years; 11 761 709 women [54.62%]). The quarterly number of participants was the highest in the last quarter of the study period (2013-2014: 4 094 951 [41.39%], 2015-2016: 4 911 629 [42.19%]); this proportion was 2.48 to 2.84 times greater than that of the first quarter. Cancer detection rates were the lowest in December (2013-2014: 0.22; 95% CI, 0.22-0.23; 2015-2016: 0.21; 95% CI, 0.21-0.22); this was approximately a 40.0% to 45.0% reduction compared with the rates in January. The age group was the significant factor for monthly detection rates. After adjustment for the age group and taking account of the number of screenings, the estimated coefficient range for the screening month was negative and the detection rate in December was significantly different than in January for both the consequent cycles (2013-2014: −0.05 to −0.18; P < .001; and 2015-2016: −0.06 to −0.19; P < .001). In the multivariable logistic model, the association of calendar month with detected cancer remained after adjusting for other confounding factors (December, 2013-2014: odds ratio, 0.82; 95% CI, 0.76-0.87; P < .001; 2015-2016: odds ratio, 0.83; 95% CI, 0.79-0.89; P < .001).

    Conclusions and Relevance  The findings of this cohort study suggest that the workload of endoscopists increased with the increasing number of examinations toward the end of the year, as demonstrated by the decreased cancer detection rates. These findings may help to improve gastric cancer detection rates of screening programs by controlling the monthly screening number and policy modifications.

    Introduction

    Although the prevalence of gastric cancer is decreasing, it remains the third leading cause of cancer mortality worldwide.1,2 Early detection, diagnosis, and treatment are the only methods to reduce mortality associated with gastric cancer. Diagnostic tools for early gastric cancer detection include upper gastrointestinal (UGI) series and endoscopy; previous studies have reported the superiority of upper endoscopy over UGI series for gastric cancer diagnosis.3-5

    Some Asian countries, including Korea, conduct endoscopic screening programs for gastric cancer.6 These screening programs may reduce the risk of death from gastric cancer.7 In Korea, the Korean National Cancer Screening Program (KNCSP) for gastric cancer was launched in 1999 to provide gastric cancer screening via upper endoscopy or UGI series for individuals aged 40 years or older.8 The total number of participants in the KNCSP for gastric cancer surveyed from 2002 to 2018 (except 2012) was 71 773 605.9-15 The individuals enrolled in the KNCSP are required to have their examinations completed by December 31 of the corresponding year or they may lose their opportunity to be examined the following year. Because most people wish to delay their examinations until the latest possible time, the number of individuals undergoing examinations might vary monthly. Furthermore, because endoscopic screening has better diagnostic accuracy than UGI series,3,4 the number of individuals undergoing endoscopy has been increasing substantially. Consequently, endoscopists’ workloads are being steadily increased.

    Several studies have shown that operator workload affects examination quality.16-21 Especially in index colonoscopy, adenoma detection rates decrease with increasing procedural hours in an endoscopist’s workload.19 We hypothesized that an increased operator workload toward the end of the year in the KNCSP would affect the quality of the screening program for gastric cancer. Controlling or limiting the number of endoscopic examinations performed in a screening program is a modifiable factor that might help to reduce gastric cancer–related mortality. In the KNCSP, a large portion of examinations was performed at the end of the year; to date, this issue has not been well described.

    To our knowledge, this is the first study to use a large population database to assess the association between the increased number of examinations over a specified period and cancer detection rates in the KNCSP. In addition, we aimed to investigate the risk factors associated with the detection of gastric cancer at screening.

    Methods
    Study Design and Population

    We conducted a retrospective, large, population-based cohort study between January 1, 2013, and December 31, 2016, using the KNCSP database after permission was obtained from the Ministry of Health and Welfare. Data were analyzed from November 1, 2019, to March 31, 2020. The total cohort comprised 26 765 665 men and women aged 40 years or older who participated in the screening program. The included participants were linked with the National Health Insurance Sharing Service–National Health Information Database (NHIS-NHID) to review their medical records until 2018. Participants were excluded if they had received previous gastric cancer diagnoses, had upper endoscopy performed for 2 consecutive years, and were not eligible for KNCSP. The study protocol was approved by the Ajou University Hospital Institutional Review Board and Ethics Committee and the requirement for informed patient consent was waived owing to the use of a deidentified data set for the analyses. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.

    Korean National Cancer Screening Program

    The KNCSP and Korea Central Cancer Registry provide targeted populations (eMethods in the Supplement) with screening services for 6 common cancers: stomach, liver, colorectum, breast, uterine cervix, and lung. For gastric cancer, the screening frequency is every 2 years, determined by the individual’s birth year. The KNCSP recommends biennial screening with either UGI series or upper endoscopy for men and women aged 40 years or older who are eligible for the KNCSP every other year. Participants are required to undergo a screening test during their designated year any time from January 1 to December 31. The participants to be examined in the relevant year visit the examination institute (eMethods and eTable 1 in the Supplement) and complete a cancer screening questionnaire and undergo an examination. On completion of the questionnaire, the data are coded and stored in the database. If gastric cancer is suspected upon endoscopic examination, biopsy sampling is performed. The specific screening endoscopy protocol is described in the eMethods in the Supplement. When a histologic diagnosis of gastric cancer is made, the findings are confirmed and recorded in the national cancer registry.

    Data Collection and Definitions

    The KNCSP data include a history of cancer screening, medical and family histories based on questionnaires, information from screening sites (region and hospital type) and screening providers, and cancer screening results. The data were separated based on the screening cycles to validate the consistency of the results found in the selected screening cycle: the first and second screening cycles were 2013-2014 and 2015-2016.

    In this analysis, the screening results were defined as positive if the endoscopic results were recorded as possible gastric cancer, early gastric cancer, or advanced gastric cancer or the biopsy results were recorded as low-grade dysplasia, high-grade dysplasia, suspicious gastric cancer, or gastric cancer. Subsequently, we linked medical records from the NHIS-NHID until December 2018 and investigated medical records to detect interval cancer. Using medical records in the NHIS-NHID, detected cancer was defined as patients with positive screening results confirmed with a diagnosis code of gastric cancer (International Statistical Classification of Diseases, Tenth Revision, code C16.xx). Interval cancer was defined as participants with negative screening results who received at least 1 diagnosis code of gastric cancer (C16.xx) in the primary diagnosis within 1 year of the negative screening result.

    Statistical Analysis

    We summarized the demographic characteristics of participants and the screening time and places using descriptive statistics and inspected trends in the number of participants based on month. Overall sensitivity, specificity, positive predictive value, detection rates per 100 positive screenings, interval cancer rates per 1000 negative screenings, and positive rates per 1000 screenings for each screening cycle were calculated. To explore monthly detection rates, we first investigated whether the number of participants varied from January to December and then calculated monthly detection rates, further stratifying the data based on age grouping.

    To identify overall trends in detection rates, we applied a Poisson regression model first; however, the overdispersion problem was encountered. Thus, a negative binomial regression model was fit for the rate outcomes to investigate whether there was any association between the detection rates and screening month. Stratified analyses taking into account the age grouping were also performed. The number of screenings per month was included as an offset term and adjusted in the model. Overdispersion was tested using scaled Pearson χ2 statistics. Moreover, we evaluated the risk factors for cancer detection using a logistic regression. The covariance matrix was multiplied by a factor of deviance/degree of freedom. The univariate logistic regression model was first applied to choose the appropriate model for each explanatory variable: the history of endoscopy; sex, age group, screening month, hospital type, and metropolitan area; and history of atrophic gastritis, ulcer, intestinal metaplasia, gastric polyp, and other gastric diseases. Then, using the stepwise selection process at a significance level of .05 for the χ2 value for entering a variable into the model and a significance level of .05 for the Wald χ2 value for a variable to stay in the model, a multivariable logistic regression analysis was performed to assess the adjusted risk factors. Detailed statistical methods are provided in eAppendix 1 and eAppendix 2 in the Supplement. Odds ratios and 95% CIs for cancer detection were obtained. All reported P values were 2-sided, and P < .05 was considered statistically significant. Analyses were performed using SAS, version 9.4 (SAS Institute Inc).

    Results
    Study Population

    Overall, 26 765 665 participants underwent gastric cancer screening; of these, 12 508 645 individuals (46.73%) underwent screening in 2013-2014 and 14 257 020 individuals (53.27%) underwent screening in 2015-2016. The overall mean (SD) age was 55.61 (10.61) years and 11 761 709 participants (54.62%) were women. A total of 9 892 812 participants for 2013-2014 were selected for inclusion in this analysis, and 11 642 410 were selected for 2015-2016. A total of 766 66 participants (0.77%) and 78 669 participants (0.68%) for each cycle screened positive. Of participants who had negative results based on the upper endoscopy, 7276 participants (0.07%) and 4531 participants (0.04%) for each respective cycle were diagnosed with gastric cancer within 1 year of endoscopic examination. Among those with positive screening results, gastric cancer was diagnosed in 28 746 individuals (37.49%) in 2013-2014 and in 31 266 individuals (39.74%) in 2015-2016 (Figure 1). For the quarterly screening periods, 41.39% individuals (4 094 951) and 42.19% individuals (4 911 629) for each respective cycle underwent endoscopic examinations during the fourth quarter (October-December) (Table 1). The values for the fourth quarter were 2.48 to 2.84 times greater than the number of examinations performed during the first quarter (2013-2014: 1 440 347 [14.56%]; 2015-2016: 1 983 392 [17.03%]). The overall screening performance for gastric cancer detection is reported in eTable 2 in the Supplement.

    Changes in the Number of Participants Screened Monthly

    The number of screenings increased from October onward (Table 1). The greatest number of participants were screened in December. For the 2013-2014 cycle, the mean number of participants between January and November was 718 502. In December, the number of participants was 2.77 times greater than the average number screened in the previous 11 months (January-November), and this number increased by 73.56% compared with the number of screenings in November. We observed a similar trend for the number of screenings performed in December for the 2015-2016 cycle (average number of participants between January and November: 849 986; 2.70 times increase in December and 57.57% increase compared with November alone). For the 2013-2014 cycle, the proportions of participants aged 40 to 59 years increased in December (January: 56.17%, December: 76.22%); however, the proportions of participants aged 60 years or older decreased from 43.83% in January to 23.78% in December. These results were similar for the 2015-2016 cycle (age 40-59 years: 54.71% in January vs 75.13% in December; ≥60 years: 45.29% in January vs 24.87% in December) (Figure 2A).

    Monthly Detection Rates for Gastric Cancer

    Overall cancer detection rates (2013-2014 and 2014-2015 cycles) showed a tendency to decrease later in the year. The detection rates were the lowest in December (2013-2014: 0.22; 95% CI, 0.22-0.23; 2015-2016: 0.21; 95% CI, 0.21-0.22) when the number of participants screened was the highest; this change was approximately a 40.0% to 45.0% decrease in detection rates compared with the rates in January (Figure 2B). For the 2013-2014 and 2015-2016 cycles, the detection rates decreased with time in the year, especially in December by age group (eFigure in the Supplement). Participants of all ages, with the exception of those in their 60s and 70s, exhibited an approximate 17.83% to 38.46% decrease in detection rates in December compared with January (eTable 3 in the Supplement). The detection rates based on endoscopic history, sex, hospital type, and screening site are reported in eTable 4 in the Supplement.

    Association Between Detection Rates and Calendar Month

    Overdispersion was tested, and nonsignificant P values (2013-2014: P = .45, 2015-2016: P = .46) suggested that the negative binomial model was a good fit for the data. Because the proportion of participants based on age group differed from January to December, the age group was added as a covariate in the model along with the month. The age group was a significant factor for both cycles. After adjustment for the age group and taking account of the number of screenings, the estimated coefficient range for the screening month was negative and the detection rate in December was significantly different than in January for both the consequent cycles (2013-2014: −0.05 to −0.18; P < .001; and 2015-2016: −0.06 to −0.19; P < .001). The detection rates for the older age group were greater than those for the younger age group (Table 2).

    Factors Associated With Endoscopic Screening Sensitivity

    In the univariate analyses, history of endoscopic examination, sex, age group, calendar month, hospital type, metropolitan area, and history of gastric diseases, including atrophic gastritis, ulcer, intestinal metaplasia, gastric polyp, and other gastric diseases, were significantly associated with gastric cancer detection (P < .001 for all) (Table 3). In the multivariable logistic regression analysis with selected indicators by the stepwise selection method, these factors remained significant, except hospital type in the 2013-2014 cycle. The overall significance of the associations of the covariates with cancer detection was similar for both cycles. We also found that the screening month was significantly associated with cancer detection (P < .001). Compared with January, the odds of cancer detection significantly decreased toward the end of the year for both KNCSP cycles.

    Discussion

    This study assessed the association between gastric cancer detection rates and calendar month using data collected from the KNCSP. The findings suggest that there was a greater number of upper endoscopy screening examinations performed during the last quarter of the year, especially in December. Most participants examined in December were aged 40 to 59 years. Gastric cancer detection rates were the lowest in December, although the greatest numbers of individuals were examined in this month. These findings were consistent regardless of age grouping and calendar cycle. We also noted that detection rates were negatively and significantly associated with calendar month. Therefore, the present findings suggest that the rapid increase in the number of screening examinations might increase endoscopists’ workloads and consequently reduce cancer detection rates.

    Upper endoscopy is the most effective examination for early diagnosis of gastric cancer.22,23 More specifically, endoscopy for screening is associated with lowering gastric cancer–related mortality.3 Furthermore, because gastric cancer prevalence increases with age,24 endoscopic evaluation as a cancer screening tool is becoming more important. Therefore, the demand for upper endoscopy is rapidly increasing the workload of the limited numbers of endoscopists. Several studies have reported that physician fatigue from being overworked affects their performance.25-29 Nevertheless, these studies have limitations, including the small patient numbers, single-center bases, and the limited number of physicians performing endoscopy. Our study used extensive data of approximately 20 million participants from the KNCSP; therefore, the outcomes have important implications.

    In the KNCSP, the number of young participants increased sharply toward December. In December, participants aged 40 to 59 years accounted for 76.22% of the included participants in the 2013-2014 cycle and 75.13% of the included participants in the 2015-2016 cycle, which was 8.84 and 6.41 times higher than the frequencies of such individuals in January. Although it is difficult to clearly explain the reason for this finding, we surmise that there may be a psychological factor associated with participants delaying the screening dates as much as possible. In addition, we think that holidays do not have an effect on the monthly variation in screening because the first and the second halves of the year have no major variations in the number of holidays in Korea. Because there was a higher number of participants with a lower gastric cancer prevalence, the decreasing rate of gastric cancer detection toward December may be natural. However, we divided the participants into 5 age groups to determine whether these findings could be explained by an increased number of younger participants. Our findings showed that detection rates decreased toward the end of the year, regardless of age group. Accordingly, we suggest that the decrease in gastric cancer detection rates toward the end of the year could not be attributed to an increased number of younger participants with a relatively lower gastric cancer prevalence.

    To reduce the selection bias due to the year of selection, we added the 2 most recent consecutive cycles from the NHIS-NHID; however, the trends were consistent regardless of cycle. Meanwhile, when the overall screening performance was investigated for the 2015-2016 cycle (eTable 2 in the Supplement), the detection rates decreased, sensitivity increased, and interval cancer rates decreased when compared with the 2013-2014 cycle. In Korea, the Ministry of Health and Welfare implemented the endoscopy quality management project of the KNCSP from 2008 to 2014. The 2012-2014 cycle was the second project cycle and, based on the results from the first project cycle, written and on-site evaluations were conducted to improve endoscopy quality management. Since 2016, the third national cancer management project has been in progress. In the KNCSP for gastric cancer, the increase in sensitivity in 2015-2016 can be considered as a positive result owing to the endoscopic quality management policy background. However, despite policies to improve the quality of endoscopic examinations, the detection rates decreased at the end of the year and the proportion of people in the young age group increased in the last quarter of the year for both cycles.

    In our study, living in the metropolitan area appeared to be associated with a reduced risk of gastric cancer detection. We assumed that gastric cancer prevalence is lower in this region owing to a high proportion of young individuals. Participants older than 80 years received more endoscopies in the nonmetropolitan area (57.0%) compared with those aged 40 to 49 years (49.0%). If further analysis were possible using cancer information and participants factor, a clearer reason might be found.

    Limitations

    This study has several limitations. First, aside from the participants’ gastric history, we did not analyze the association between participant-level factors and gastric cancer detection. Participant factors, including symptoms, may influence the timing of the examination; however, we could not confirm this from the KNCSP’s available data. We could not obtain the information on the participants’ comorbidities; therefore, it could not be included in the logistic regression analysis. Second, this study was based on data from 2013 to 2016—these were the latest data we could access. Although this was a study containing real-world data for more than 20 million individuals with confirmation of cancer occurrence and follow-up records, our findings may not be sufficiently representative. Additional studies assessing the entire cohort of individuals who have undergone subsequent screening may clarify the study outcomes. Third, an objective evaluation of the fatigue experienced by endoscopists was not performed. Fourth, cancer stages and their possible effects were not analyzed in this study owing to data unavailability. Fifth, because this study used screening program data, there may be selection bias.

    Conclusions

    We used national population-based data from the KNCSP database to observe a decrease in the gastric cancer detection rates toward the end of the year as the number of examinations rapidly increased. Our findings suggest that policy guidelines are needed to support improvements to screening programs for gastric cancer detection using endoscopy. Such adapted guidelines could contain recommendations on maintaining limits of the number of examinations allowed within a specific period. Further studies should be performed to determine the development index of operator workload or burnout in professionals performing UGI endoscopy to assist in establishing guidelines highlighting the best performance of each operator.

    Back to top
    Article Information

    Accepted for Publication: November 11, 2020.

    Published: January 7, 2021. doi:10.1001/jamanetworkopen.2020.32542

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Noh C-K et al. JAMA Network Open.

    Corresponding Author: Kee Myung Lee, MD, PhD, Department of Gastroenterology, Ajou University School of Medicine, 164 World Cup-ro, Yeongtong-gu, Suwon 16499, Republic of Korea (lkm5104@ajou.ac.kr).

    Author Contributions: Dr K. M. Lee had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Dr Noh and Ms E. Lee contributed equally to the study.

    Concept and design: Noh, E. Lee, Cheong, K. M. Lee.

    Acquisition, analysis, or interpretation of data: Noh, E. Lee, G. H. Lee, Kang, Lim, B. Park, J. B. Park, Shin, Kim, K. M. Lee.

    Drafting of the manuscript: Noh, E. Lee, B. Park, Cheong, K. M. Lee.

    Critical revision of the manuscript for important intellectual content: Noh, E. Lee, G. H. Lee, Kang, Lim, J. B. Park, Shin, Kim, K. M. Lee.

    Statistical analysis: E. Lee, B. Park.

    Administrative, technical, or material support: Noh, E. Lee, G. H. Lee.

    Supervision: E. Lee, Cheong, Kim.

    Conflict of Interest Disclosures: None reported.

    Additional Contributions: Woohyun Cho, MFA (Medical Information & Media Center, Ajou University School of Medicine), provided editing services for images and illustrations. She did not receive compensation.

    References
    1.
    Ferlay  J, Soerjomataram  I, Dikshit  R,  et al.  Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012.   Int J Cancer. 2015;136(5):E359-E386. doi:10.1002/ijc.29210 PubMedGoogle ScholarCrossref
    2.
    Bray  F, Ferlay  J, Soerjomataram  I, Siegel  RL, Torre  LA, Jemal  A.  Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries.   CA Cancer J Clin. 2018;68(6):394-424. doi:10.3322/caac.21492 PubMedGoogle ScholarCrossref
    3.
    Jun  JK, Choi  KS, Lee  HY,  et al.  Effectiveness of the Korean National Cancer Screening Program in reducing gastric cancer mortality.   Gastroenterology. 2017;152(6):1319-1328.e7. doi:10.1053/j.gastro.2017.01.029 PubMedGoogle ScholarCrossref
    4.
    Tashiro  A, Sano  M, Kinameri  K, Fujita  K, Takeuchi  Y.  Comparing mass screening techniques for gastric cancer in Japan.   World J Gastroenterol. 2006;12(30):4873-4874. doi:10.3748/wjg.v12.i30.4873PubMedGoogle Scholar
    5.
    Lee  HY, Park  EC, Jun  JK, Choi  KS, Hahm  MI.  Comparing upper gastrointestinal X-ray and endoscopy for gastric cancer diagnosis in Korea.   World J Gastroenterol. 2010;16(2):245-250. doi:10.3748/wjg.v16.i2.245 PubMedGoogle ScholarCrossref
    6.
    Leung  WK, Wu  MS, Kakugawa  Y,  et al; Asia Pacific Working Group on Gastric Cancer.  Screening for gastric cancer in Asia: current evidence and practice.   Lancet Oncol. 2008;9(3):279-287. doi:10.1016/S1470-2045(08)70072-X PubMedGoogle ScholarCrossref
    7.
    Zhang  X, Li  M, Chen  S,  et al.  Endoscopic screening in Asian countries is associated with reduced gastric cancer mortality: a meta-analysis and systematic review.   Gastroenterology. 2018;155(2):347-354.e9. doi:10.1053/j.gastro.2018.04.026 PubMedGoogle ScholarCrossref
    8.
    Shin  HR, Won  YJ, Jung  KW,  et al; Members of the Regional Cancer Registries.  Nationwide cancer incidence in Korea, 1999-2001; first result using the national cancer incidence database.   Cancer Res Treat. 2005;37(6):325-331. doi:10.4143/crt.2005.37.6.325 PubMedGoogle ScholarCrossref
    9.
    Lee  S, Jun  JK, Suh  M,  et al.  Gastric cancer screening uptake trends in Korea: results for the National Cancer Screening Program from 2002 to 2011: a prospective cross-sectional study.   Medicine (Baltimore). 2015;94(8):e533. doi:10.1097/MD.0000000000000533 PubMedGoogle Scholar
    10.
    National Health Insurance Service.  2013 National Health Screening Statistical Yearbook. National Health Insurance Service; 2014.
    11.
    National Health Insurance Service.  2014 National Health Screening Statistical Yearbook. National Health Insurance Service; 2015.
    12.
    National Health Insurance Service.  2015 National Health Screening Statistical Yearbook. National Health Insurance Service; 2016.
    13.
    National Health Insurance Service.  2016 National Health Screening Statistical Yearbook. National Health Insurance Service; 2017.
    14.
    National Health Insurance Service.  2017 National Health Screening Statistical Yearbook. National Health Insurance Service; 2018.
    15.
    National Health Insurance Service.  2018 National Health Screening Statistical Yearbook. National Health Insurance Service; 2019.
    16.
    Harewood  GC, Chrysostomou  K, Himy  N, Leong  WL.  Impact of operator fatigue on endoscopy performance: implications for procedure scheduling.   Dig Dis Sci. 2009;54(8):1656-1661. doi:10.1007/s10620-008-0549-7 PubMedGoogle ScholarCrossref
    17.
    Kaneshiro  M, Ho  A, Chan  M, Cohen  H, Spiegel  BM.  Colonoscopy yields fewer polyps as the day progresses despite using social influence theory to reverse the trend.   Gastrointest Endosc. 2010;72(6):1233-1240. doi:10.1016/j.gie.2010.08.034 PubMedGoogle ScholarCrossref
    18.
    Chan  MY, Cohen  H, Spiegel  BM.  Fewer polyps detected by colonoscopy as the day progresses at a Veteran’s Administration teaching hospital.   Clin Gastroenterol Hepatol. 2009;7(11):1217-1223. doi:10.1016/j.cgh.2009.07.013 PubMedGoogle ScholarCrossref
    19.
    Almadi  MA, Sewitch  M, Barkun  AN, Martel  M, Joseph  L.  Adenoma detection rates decline with increasing procedural hours in an endoscopist’s workload.   Can J Gastroenterol Hepatol. 2015;29(6):304-308. doi:10.1155/2015/789038 PubMedGoogle ScholarCrossref
    20.
    Rex  DK.  Quality in colonoscopy: cecal intubation first, then what?   Am J Gastroenterol. 2006;101(4):732-734. doi:10.1111/j.1572-0241.2006.00483.x PubMedGoogle ScholarCrossref
    21.
    Lee  A, Jensen  CD, Marks  AR,  et al.  Endoscopist fatigue estimates and colonoscopic adenoma detection in a large community-based setting.   Gastrointest Endosc. 2017;85(3):601-610.e2. doi:10.1016/j.gie.2016.09.033 PubMedGoogle ScholarCrossref
    22.
    Hamashima  C.  Current issues and future perspectives of gastric cancer screening.   World J Gastroenterol. 2014;20(38):13767-13774. doi:10.3748/wjg.v20.i38.13767 PubMedGoogle ScholarCrossref
    23.
    Choi  KS, Jun  JK, Park  EC,  et al.  Performance of different gastric cancer screening methods in Korea: a population-based study.   PLoS One. 2012;7(11):e50041. doi:10.1371/journal.pone.0050041 PubMedGoogle Scholar
    24.
    Karimi  P, Islami  F, Anandasabapathy  S, Freedman  ND, Kamangar  F.  Gastric cancer: descriptive epidemiology, risk factors, screening, and prevention.   Cancer Epidemiol Biomarkers Prev. 2014;23(5):700-713. doi:10.1158/1055-9965.EPI-13-1057 PubMedGoogle ScholarCrossref
    25.
    Saposnik  G, Baibergenova  A, Bayer  N, Hachinski  V.  Weekends: a dangerous time for having a stroke?   Stroke. 2007;38(4):1211-1215. doi:10.1161/01.STR.0000259622.78616.ea PubMedGoogle ScholarCrossref
    26.
    Bendavid  E, Kaganova  Y, Needleman  J, Gruenberg  L, Weissman  JS.  Complication rates on weekends and weekdays in US hospitals.   Am J Med. 2007;120(5):422-428. doi:10.1016/j.amjmed.2006.05.067 PubMedGoogle ScholarCrossref
    27.
    Hendey  GW, Barth  BE, Soliz  T.  Overnight and postcall errors in medication orders.   Acad Emerg Med. 2005;12(7):629-634. doi:10.1197/j.aem.2005.02.009 PubMedGoogle ScholarCrossref
    28.
    Maung  R.  Pathologists' workload and patient safety.   Diagn Histopathol. 2016;22(8):283-287. doi:10.1016/j.mpdhp.2016.07.004 Google ScholarCrossref
    29.
    Sanaka  MR, Shah  N, Mullen  KD, Ferguson  DR, Thomas  C, McCullough  AJ.  Afternoon colonoscopies have higher failure rates than morning colonoscopies.   Am J Gastroenterol. 2006;101(12):2726-2730. doi:10.1111/j.1572-0241.2006.00887.x PubMedGoogle ScholarCrossref
    ×