Mapping the incidence rate of typhoid fever in sub-Saharan Africa.

BACKGROUND: With more than 1.2 million illnesses and 29,000 deaths in sub-Saharan Africa in 2017, typhoid fever continues to be a major public health problem. Effective control of the disease would benefit from an understanding of the subnational geospatial distribution of the disease incidence. METHOD: We collated records of the incidence rate of typhoid fever confirmed by culture of blood in Africa from 2000 to 2022. We estimated the typhoid incidence rate for sub-Saharan Africa on 20 km × 20 km grids by exploring the association with geospatial covariates representing access to improved water and sanitation, health conditions of the population, and environmental conditions. RESULTS: We identified six published articles and one pre-print representing incidence rate estimates in 22 sites in 2000-2022. Estimated incidence rates showed geospatial variation at sub-national, national, and regional levels. The incidence rate was high in Western and Eastern African subregions followed by Southern and Middle African subregions. By age, the incidence rate was highest among 5-14 yo followed by 2-4 yo, > 14 yo, and 0-1 yo. When aggregated across all age classes and grids that comprise each country, predicted incidence rates ranged from 43.7 (95% confidence interval: 0.6 to 591.2) in Zimbabwe to 2,957.8 (95% CI: 20.8 to 4,245.2) in South Sudan per 100,000 person-years. Sub-national heterogeneity was evident with the coefficient of variation at the 20 km × 20 km grid-level ranging from 0.7 to 3.3 and was generally lower in high-incidence countries and widely varying in low-incidence countries. CONCLUSION: Our study provides estimates of 20 km × 20 km incidence rate of typhoid fever across sub-Saharan Africa based on data collected from 2000 through 2020. Increased understanding of the subnational geospatial variation of typhoid fever in Africa may inform more effective intervention programs by better targeting resources to heterogeneously disturbed disease risk.

Effect of characteristics on the clinical course at the initiation of treatment for human immunodeficiency virus infection using dimensionality reduction.

The beginning of human immunodeficiency virus (HIV) infection treatment depends on various factors, which are significantly correlated with the initial CD4 cell number. However, a covariate correlation between these factors may not reflect the correct outcome variable. Thus, we evaluated the effects of a combination of fixed factors (reduced dimensions), which determine when to start treatment for the first time, on short-term outcome, long-term outcome, and survival, considering correlations between factors. Multiple correspondence analysis was performed on variables obtained from 925 patients who participated in a Korean HIV/acquired immunodeficiency syndrome cohort study (2006-2017). Five reduced dimension groups were derived according to clinical data, viral load, CD4 cell count at diagnosis, initial antiretroviral therapy, and others. The dimension group with high initial viral loads (55,000 copies/mL) and low CD4 cell counts (< 200 cells/mm3) should start treatment promptly after diagnosis. Groups with high initial CD4 cell counts (> 350 cells/mm3) that did not require immediate treatment according to previous guidelines had a higher failure rate for long-term relative CD4 recovery. Our results highlight the importance of early diagnosis and treatment to positively influence long-term disease outcomes, even if the initial immune status is poor, given the patient's combination of early diagnostic symptoms.

An analysis of the dynamic spatial spread of COVID-19 across South Korea.

The first case of coronavirus disease 2019 (COVID-19) in South Korea was confirmed on January 20, 2020, approximately three weeks after the report of the first COVID-19 case in Wuhan, China. By September 15, 2021, the number of cases in South Korea had increased to 277,989. Thus, it is important to better understand geographical transmission and design effective local-level pandemic plans across the country over the long term. We conducted a spatiotemporal analysis of weekly COVID-19 cases in South Korea from February 1, 2020, to May 30, 2021, in each administrative region. For the spatial domain, we first covered the entire country and then focused on metropolitan areas, including Seoul, Gyeonggi-do, and Incheon. Moran's I and spatial scan statistics were used for spatial analysis. The temporal variation and dynamics of COVID-19 cases were investigated with various statistical visualization methods. We found time-varying clusters of COVID-19 in South Korea using a range of statistical methods. In the early stage, the spatial hotspots were focused in Daegu and Gyeongsangbuk-do. Then, metropolitan areas were detected as hotspots in December 2020. In our study, we conducted a time-varying spatial analysis of COVID-19 across the entirety of South Korea over a long-term period and found a powerful approach to demonstrating the current dynamics of spatial clustering and understanding the dynamic effects of policies on COVID-19 across South Korea. Additionally, the proposed spatiotemporal methods are very useful for understanding the spatial dynamics of COVID-19 in South Korea.

Investigating the spatio-temporal variation of hepatitis A in Korea using a Bayesian model.

Hepatitis A is a water-borne infectious disease that frequently occurs in unsanitary environments. However, paradoxically, those who have spent their infancy in a sanitary environment are more susceptible to hepatitis A because they do not have the opportunity to acquire natural immunity. In Korea, hepatitis A is prevalent because of the distribution of uncooked seafood, especially during hot and humid summers. In general, the transmission of hepatitis A is known to be dynamically affected by socioeconomic, environmental, and weather-related factors and is heterogeneous in time and space. In this study, we aimed to investigate the spatio-temporal variation of hepatitis A and the effects of socioeconomic and weather-related factors in Korea using a flexible spatio-temporal model. We propose a Bayesian Poisson regression model coupled with spatio-temporal variability to estimate the effects of risk factors. We used weekly hepatitis A incidence data across 250 districts in Korea from 2016 to 2019. We found spatial and temporal autocorrelations of hepatitis A indicating that the spatial distribution of hepatitis A varied dynamically over time. From the estimation results, we noticed that the districts with large proportions of males and foreigners correspond to higher incidences. The average temperature was positively correlated with the incidence, which is in agreement with other studies showing that the incidences in Korea are noticeable in spring and summer due to the increased outdoor activity and intake of stale seafood. To the best of our knowledge, this study is the first to suggest a spatio-temporal model for hepatitis A across the entirety of Korean. The proposed model could be useful for predicting, preventing, and controlling the spread of hepatitis A.

Objectively Measured Built Environments and Cardiovascular Diseases in Middle-Aged and Older Korean Adults.

This study assesses the association between the objectively measured built environment and cardiovascular diseases (CVDs) in 50,741 adults from the Korean Community Health Survey. The CVD outcomes of hypertension, diabetes, dyslipidemia, stroke, and myocardial infarction (MI) or angina were derived from self-reported histories of physician diagnoses. Using ArcGIS software and Korean government databases, this study measured the built environment variables for the 546 administrative areas of Gyeonggi province. A Bayesian spatial multilevel model was performed independently in two age groups (i.e., 40-59 years or ≥60 years). After adjusting for statistical significant individual- and community-level factors with the spatial associations, living far from public transit was associated with an increase in the odds of MI or angina in middle-aged adults, while living in neighborhoods in which fast-food restaurants were concentrated was associated with a decrease in the odds of hypertension and stroke. For adults 60 or older, living farther from public physical-activity (PA) facilities was associated with a 15% increased odds for dyslipidemia, compared with living in neighborhoods nearer to PA facilities. These findings suggest that creating a built environment that provides more opportunities to engage in PA in everyday life should be considered a strategy to reduce the prevalence of CVD.

Time-variant reproductive number of COVID-19 in Seoul, Korea.

OBJECTIVES: To estimate time-variant reproductive number (Rt) of coronavirus disease 19 based on either number of daily confirmed cases or their onset date to monitor effectiveness of quarantine policies. METHODS: Using number of daily confirmed cases from January 23, 2020 to March 22, 2020 and their symptom onset date from the official website of the Seoul Metropolitan Government and the district office, we calculated Rt using program R's package "EpiEstim". For asymptomatic cases, their symptom onset date was considered as -2, -1, 0, +1, and +2 days of confirmed date. RESULTS: Based on the information of 313 confirmed cases, the epidemic curve was shaped like 'propagated epidemic curve'. The daily Rt based on Rt_c peaked to 2.6 on February 20, 2020, then showed decreased trend and became <1.0 from March 3, 2020. Comparing both Rt from Rt_c and from the number of daily onset cases, we found that the pattern of changes was similar, although the variation of Rt was greater when using Rt_c. When we changed assumed onset date for asymptotic cases (-2 days to +2 days of the confirmed date), the results were comparable. CONCLUSIONS: Rt can be estimated based on Rt_c which is available from daily report of the Korea Centers for Disease Control and Prevention. Estimation of Rt would be useful to continuously monitor the effectiveness of the quarantine policy at the city and province levels.

Spatial epidemic dynamics of the COVID-19 outbreak in China.

BACKGROUND: On 31 December 2019 an outbreak of COVID-19 in Wuhan, China, was reported. The outbreak spread rapidly to other Chinese cities and multiple countries. This study described the spatio-temporal pattern and measured the spatial association of the early stages of the COVID-19 epidemic in mainland China from 16 January-06 February 2020. METHODS: This study explored the spatial epidemic dynamics of COVID-19 in mainland China. Moran's I spatial statistic with various definitions of neighbours was used to conduct a test to determine whether a spatial association of the COVID-19 infections existed. RESULTS: The spatial spread of the COVID-19 pandemic in China was observed. The results showed that most of the models, except medical-care-based connection models, indicated a significant spatial association of COVID-19 infections from around 22 January 2020. CONCLUSIONS: Spatial analysis is of great help in understanding the spread of infectious diseases, and spatial association was the key to the spatial spread during the early stages of the COVID-19 pandemic in mainland China.

Implementation of Complementary Model using Optimal Combination of Hematological Parameters for Sepsis Screening in Patients with Fever.

The early detection and timely treatment are the most important factors for improving the outcome of patients with sepsis. Sepsis-related clinical score, such as SIRS, SOFA and LODS, were defined to identify patients with suspected infection and to predict severity and mortality. A few hematological parameters associated with organ dysfunction and infection were included in the score although various clinical pathology parameters (hematology, serum chemistry and plasma coagulation) in blood sample have been found to be associated with outcome in patients with sepsis. The investigation of the parameters facilitates the implementation of a complementary model for screening sepsis to existing sepsis clinical criteria and other laboratory signs. In this study, statistical analysis on the multiple clinical pathology parameters obtained from two groups, patients with sepsis and patients with fever, was performed and the complementary model was elaborated by stepwise parameter selection and machine learning. The complementary model showed statistically better performance (AUC 0.86 vs. 0.74-0.51) than models built up with specific hematology parameters involved in each existing sepsis-related clinical score. Our study presents the complementary model based on the optimal combination of hematological parameters for sepsis screening in patients with fever.

Influence of Neighborhood Environment on Korean Adult Obesity Using a Bayesian Spatial Multilevel Model.

Previous studies using spatial statistical modeling that account for spatial associations between geographic areas are scarce. Therefore, this study examines the association between neighborhood environment and obesity using a Bayesian spatial multilevel model. Data from 78,014 adults living in Gyeonggi province in Korea were drawn from the 2013-2014 Korean Community Health Survey. Korean government databases and ArcGIS software (version 10.1, ESRI, Redlands, CA) were used to measure the neighborhood environment for 546 administrative districts of Gyeonggi province. A Bayesian spatial multilevel model was implemented across gender and age groups. The findings indicate that women aged 19-39 years who lived in neighborhoods farthest away from parks were more likely to be obese. Men aged 40-59 years who lived in neighborhoods farther from public physical activity facilities and with lower population density were more likely to be obese. Obesity for women aged 19-39 years was the most spatially dependent, while obesity for women aged 40-59 years was the least spatially dependent. The results suggest that neighborhood environments that provide more opportunities for physical activity are negatively related to obesity. Therefore, the creation of physical activity in favorable neighborhood environments, considering gender and age, may be a valuable strategy to reduce obesity.

Space-Time Relationship between Short-Term Exposure to Fine and Coarse Particles and Mortality in a Nationwide Analysis of Korea: A Bayesian Hierarchical Spatio-Temporal Model.

Previous studies have shown an association between mortality and ambient air pollution in South Korea. However, these studies may have been subject to bias, as they lacked adjustment for spatio-temporal structures. This paper addresses this research gap by examining the association between air pollution and cause-specific mortality in South Korea between 2012 and 2015 using a two-stage Bayesian spatio-temporal model. We used 2012-2014 mortality and air pollution data for parameter estimation (i.e., model fitting) and 2015 data for model validation. Our results suggest that the relative risks of total, cardiovascular, and respiratory mortality were 1.028, 1.047, and 1.045, respectively, with every 10-µg/m3 increase in monthly PM2.5 (fine particulate matter) exposure. These findings warrant protection of populations who experience elevated ambient air pollution exposure to mitigate mortality burden in South Korea.

A Bayesian two-stage spatially dependent variable selection model for space-time health data.

In space-time epidemiological modeling, most studies have considered the overall variations in relative risk to better estimate the effects of risk factors on health outcomes. However, the associations between risk factors and health outcomes may vary across space and time. Especially, the temporal patterns of the covariate effects may depend on space. Thus, we propose a Bayesian two-stage spatially dependent variable selection approach for space-time health data to determine the spatially varying subsets of regression coefficients with common temporal dependence. The two-stage structure allows reduction of the spatial confounding bias in the estimates of the regression coefficients. A simulation study is conducted to examine the performance of the proposed two-stage model. We apply the proposed model to the number of inpatients with lung cancer in 159 counties of Georgia, USA.

Bayesian zero-inflated spatio-temporal modelling of scrub typhus data in Korea, 2010-2014.

Scrub typhus, a bacterial, febrile disease commonly occurring in the autumn, can easily be cured if diagnosed early. However, it can develop serious complications and even lead to death. For this reason, it is an important issue to find the risk factors and thus be able to prevent outbreaks. We analyzed the monthly scrub typhus data over the entire areas of South Korea from 2010 through 2014. A 2-stage hierarchical framework was considered since weather data are covariates and the scrub typhus data have different spatial resolutions. At the first stage, we obtained the administrative-level estimates for weather data using a spatial model; in the second, we applied a Bayesian zero-inflated spatio-temporal model since the scrub typhus data include excess zero counts. We found that the zero-inflated model considering the spatio-temporal interaction terms improves fitting and prediction performance. This study found that low humidity and a high proportion of elderly people are significantly associated with scrub typhus incidence.

Transcriptome Analysis Reveals Nonfoamy Rather Than Foamy Plaque Macrophages Are Proinflammatory in Atherosclerotic Murine Models.

RATIONALE: Monocyte infiltration into the subintimal space and its intracellular lipid accumulation are the most prominent features of atherosclerosis. To understand the pathophysiology of atherosclerotic disease, we need to understand the characteristics of lipid-laden foamy macrophages in the subintimal space during atherosclerosis. OBJECTIVE: We sought to examine the transcriptomic profiles of foamy and nonfoamy macrophages isolated from atherosclerotic intima. METHODS AND RESULTS: Single-cell RNA sequencing analysis of CD45+ leukocytes from murine atherosclerotic aorta revealed that there are macrophage subpopulations with distinct differentially expressed genes involved in various functional pathways. To specifically characterize the intimal foamy macrophages of plaque, we developed a lipid staining-based flow cytometric method for analyzing the lipid-laden foam cells of atherosclerotic aortas. We used the fluorescent lipid probe BODIPY493/503 and assessed side-scattered light as an indication of cellular granularity. BODIPYhiSSChi foamy macrophages were found residing in intima and expressing CD11c. Foamy macrophage accumulation determined by flow cytometry was positively correlated with the severity of atherosclerosis. Bulk RNA sequencing analysis showed that compared with nonfoamy macrophages, foamy macrophages expressed few inflammatory genes but many lipid-processing genes. Intimal nonfoamy macrophages formed the major population expressing IL (interleukin)-1ß and many other inflammatory transcripts in atherosclerotic aorta. CONCLUSIONS: RNA sequencing analysis of intimal macrophages from atherosclerotic aorta revealed that lipid-loaded plaque macrophages are not likely the plaque macrophages that drive lesional inflammation.

Poor Oral Health Predicts Higher 10-Year Cardiovascular Risk: A Propensity Score Matching Analysis.

BACKGROUND: Although the association between periodontitis and cardiovascular disease (CVD) has been reported, whether periodontitis and poor oral health behavior influence CVD risk remains unclear. OBJECTIVE: The aim of this study was to examine whether periodontal disease and poor oral health behavior predict 10-year general CVD risk using the Framingham Risk Score. METHODS: Patients older than 30 years with no CVD history (n = 8370) were selected using cross-sectional study data from the Korean National Health and Nutrition Examination Survey in 2013 and 2014. To reduce selection bias in this population-based study, propensity score matching analysis was used with SPSS and R programs to compare CVD risk. RESULTS: Overall, 39.2% of the study population (n = 3277) had a global CVD risk of  10% or greater. In the low- and at-risk groups, 20.7% and 45.3% of patients, respectively, were diagnosed with periodontal disease by a dentist. Moreover, 43.2% and 62.8% of the low- and at-risk group patients, respectively, brushed teeth less than 3 times a day. After 1:1 propensity score matching of the low-risk (n = 1135) and at-risk (n = 1135) groups, bivariate analyses showed that a diagnosis of periodontal disease and less frequent toothbrushing were associated with a higher CVD risk (P < .001). Logistic regression analysis also showed that patients having periodontal disease and who brushed teeth less frequently were 1.38 and 1.33 times, respectively, more likely to be at risk of CVD (P < .001). CONCLUSION: Education on periodontitis management and oral hygiene behavior should be included, when strategies for public risk reduction of CVD are developed.

Bayesian spatially dependent variable selection for small area health modeling.

Statistical methods for spatial health data to identify the significant covariates associated with the health outcomes are of critical importance. Most studies have developed variable selection approaches in which the covariates included appear within the spatial domain and their effects are fixed across space. However, the impact of covariates on health outcomes may change across space and ignoring this behavior in spatial epidemiology may cause the wrong interpretation of the relations. Thus, the development of a statistical framework for spatial variable selection is important to allow for the estimation of the space-varying patterns of covariate effects as well as the early detection of disease over space. In this paper, we develop flexible spatial variable selection approaches to find the spatially-varying subsets of covariates with significant effects. A Bayesian hierarchical latent model framework is applied to account for spatially-varying covariate effects. We present a simulation example to examine the performance of the proposed models with the competing models. We apply our models to a county-level low birth weight incidence dataset in Georgia.

Breast Cancer Detection in a Screening Population: Comparison of Digital Mammography, Computer-Aided Detection Applied to Digital Mammography and Breast Ultrasound.

PURPOSE: We aimed to compare the detection of breast cancer using full-field digital mammography (FFDM), FFDM with computer-aided detection (FFDM+CAD), ultrasound (US), and FFDM+CAD plus US (FFDM+CAD+US), and to investigate the factors affecting cancer detection. METHODS: In this retrospective study conducted from 2008 to 2012, 48,251 women underwent FFDM and US for cancer screening. One hundred seventy-one breast cancers were detected: 115 invasive cancers and 56 carcinomas in situ. Two radiologists evaluated the imaging findings of FFDM, FFDM+CAD, and US, based on the Breast Imaging Reporting and Data System lexicon of the American College of Radiology by consensus. We reviewed the clinical and the pathological data to investigate factors affecting cancer detection. We statistically used generalized estimation equations with a logit link to compare the cancer detectability of different imaging modalities. To compare the various factors affecting detection versus nondetection, we used Wilcoxon rank sum, chi-square, or Fisher exact test. RESULTS: The detectability of breast cancer by US (96.5%) or FFDM+CAD+US (100%) was superior to that of FFDM (87.1%) (p=0.019 or p<0.001, respectively) or FFDM+ CAD (88.3%) (p=0.050 or p<0.001, respectively). However, cancer detectability was not significantly different between FFDM versus FFDM+CAD (p=1.000) and US alone versus FFDM+CAD+US (p=0.126). The tumor size influenced cancer detectability by all imaging modalities (p<0.050). In FFDM and FFDM+CAD, the nondetecting group consisted of younger patients and patients with a denser breast composition (p<0.050). In breast US, carcinoma in situ was more frequent in the nondetecting group (p=0.014). CONCLUSION: For breast cancer screening, breast US alone is satisfactory for all age groups, although FFDM+ CAD+US is the perfect screening method. Patient age, breast composition, and pathological tumor size and type may influence cancer detection during screening.

Prestroke Antiplatelet Agents in First-Time Ischemic Stroke Are Related to Subtypes.

BACKGROUND: It is not well known whether prestroke antiplatelet agents (PAs) are associated with the subtypes of ischemic stroke. METHODS: We screened patients in a hospital-based stroke registry. Patients who were admitted with a diagnosis of first-time ischemic stroke within 5 days of symptom onset were included. Ischemic stroke subtypes were classified in accordance with the Trial of ORG 10172 in Acute Stroke Treatment classification based on stroke mechanism: large-artery atherosclerosis (LA), cardioembolism (CE), small vessel occlusion (SVO), other determined (OC) or undetermined causes (UC). Multinomial logistic regression analyses were performed to evaluate the effect of PA on stroke subtypes before and after propensity score matching. RESULTS: Among 3,025 patients, 748 (24.7%) were taking antiplatelet agents prior to stroke. After propensity score matching, 1,190 patients were ultimately included. The PA group was associated with strokes caused by SVO rather than LA in multinomial logistic regression of an unmatched dataset. However, multivariable analysis after propensity score matching demonstrated that PA use was associated with a higher probability of SVO and CE (OR 2.05, p < 0.001 and OR 1.62, p = 0.05, respectively) compared with LA. CONCLUSIONS: PAs were associated with specific index stroke subtypes.

Prestroke antiplatelet agents in first-ever ischemic stroke: clinical effects.

OBJECTIVE: To investigate whether prestroke antiplatelet agent (PA) use was associated with initial stroke severity. METHODS: This was a retrospective, case-control study based on data from a prospectively collected hospital-based stroke registry (Korea University Stroke Registry). A total of 3,025 patients who were admitted with a diagnosis of first-ever ischemic stroke within 5 days of symptom onset were included. Stroke severity was measured with the NIH Stroke Scale (NIHSS). NIHSS score ≤4 at admission was categorized as mild stroke. Patients from the PA group were matched with those from the non-PA group using estimated propensity scores at a 1:1 ratio. Stepwise multivariable logistic regression analyses were performed on patients in the matched datasets with initial mild stroke. RESULTS: Patients' mean age was 66.3 ± 13.0 years, and 1,850 were men (61.5%). A total of 748 patients had been taking antiplatelet agents prior to stroke onset; 644 patients (86.1%) were taking a single antiplatelet agent. Among these agents, aspirin (83.7%) was the most common. A total of 102 patients (13.6%) were taking 2 antiplatelet agents. Multivariable analysis after propensity score matching demonstrated that PA use was associated with initial mild stroke (odds ratio 1.344; 95% confidence interval 1.014-1.782). CONCLUSIONS: PA use was associated with decreased first-ever stroke severity, suggesting that it has a beneficial effect. CLASSIFICATION OF EVIDENCE: This study provides Class II evidence that prestroke use of antiplatelet agents reduces stroke severity in patients with first-ever acute ischemic stroke.

Patterns of decrease in multidirectional myocardial deformations in patients with fluctuating left ventricular ejection fraction.

Few studies have examined the variations in longitudinal/circumferential/radial strain (LS/CS/RS) and strain rate (LSr/CSr/RSr) in individual hearts when the left ventricular ejection fraction (LVEF) has changed. We hypothesized the relationships of strain/strain rate and LVEF are not linear, but vary with multiple inflection points (IPs) in individual hearts.Twenty-five patients with fluctuating LVEF (ΔLVEF > 10%) who had 2-D speckle tracking echocardiography available for analysis were enrolled. After models of best fit were obtained from the 'collective' plots to determine inflection points, the decrements of slopes above inflection points (IP) were compared with those below IPs in the 'individual hearts' plots.In the 'collective' plots, both LS and LSr linearly decreased in proportion to LVEF when LVEF ≥ 40% but remained constant regardless of LVEF when LVEF < 40% (IPs when LVEF = 40%, P < 0.0001). The RS-LVEF relationship was sigmoid with two IPs when LVEF = 30% and 50% (P < 0.0001). However, in the 'individual hearts' plots, the decrements of slopes above and below IPs were not different for LS-LVEF and LSr-LVEF, and marginally different for RS-LVEF (P = 0.049, across IP when LVEF = 50%).Collectively, the relationship of LS/LSr/RS and LVEF seemed to be not linear, but inflective, however, we could not prove the inflective relationship in individual hearts with fluctuating LVEF. Further study with more patients is needed to prove our hypothesis.

Risk of adenomas with high-risk characteristics based on two previous colonoscopy.

BACKGROUND AND AIM: The recommended intervals between surveillance colonoscopies are based on the most recent examination findings. However, whether the two previous colonoscopies affect second surveillance colonoscopic findings is not established. The aim of this study is to estimate the risk of obtaining high-risk findings (HRF) on the next surveillance colonoscopy using the results of two previous colonoscopies, and to estimate the appropriate time interval for the next surveillance colonoscopy. METHODS: Among subjects who underwent screening colonoscopy during January 2002-December 2009, patients who underwent second surveillance colonoscopy before June 2012 were enrolled. "No adenoma" was defined as a hyperplastic polyp or no polyp, "low-risk findings (LRF)" as one or two small (< 1 cm) tubular adenomas, and "HRF" as advanced adenoma, cancer, or any sized multiple (≥ 3) adenomas. RESULTS: Among enrolled 852 subjects, 65 (7.6%) had HRF at second surveillance colonoscopy. Multivariate analysis showed that HRF on second surveillance colonoscopy were associated with male and HRF on screening colonoscopy (all, P < 0.01). In subjects with LRF on first surveillance colonoscopy, HRF on the screening colonoscopy significantly affected the detection of HRF on second surveillance colonoscopy (P < 0.01). Patients with HRF on screening colonoscopy and LRF on the first surveillance colonoscopy had no different risk of HRF on second surveillance colonoscopy from those with HRF on first surveillance colonoscopy (P > 0.05). CONCLUSIONS: The HRF on second surveillance are significantly associated with previous two colonoscopic results. In patients with LRF on first surveillance, screening colonoscopic findings should be considered to determine the optimal surveillance interval.