BackgroundThe importance of identifying and evaluating adverse drug reactions (ADRs) has been widely recognized. Many studies have developed algorithms for ADR signal detection using electronic health record (EHR) data. In this study, we propose a machine learning (ML) model that enables accurate ADR signal detection by integrating features from existing algorithms based on inpatient EHR laboratory results.Materials and methodsTo construct an ADR reference dataset, we extracted known drug–laboratory event pairs represented by a laboratory test from the EU-SPC and SIDER databases. All possible drug–laboratory event pairs, except known ones, are considered unknown. To detect a known drug–laboratory event pair, three existing algorithms—CERT, CLEAR, and PACE—were applied to 21-year inpatient EHR data. We also constructed ML models (based on random forest, L1 regularized logistic regression, support vector machine, and a neural network) that use the intermediate products of the CERT, CLEAR, and PACE algorithms as inputs and determine whether a drug–laboratory event pair is associated. For performance comparison, we evaluated the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), F1-measure, and area under receiver operating characteristic (AUROC).ResultsAll measures of ML models outperformed those of existing algorithms with sensitivity of 0.593–0.793, specificity of 0.619–0.796, NPV of 0.645–0.727, PPV of 0.680–0.777, F1-measure of 0.629–0.709, and AUROC of 0.737–0.816. Features related to change or distribution of shape were considered important for detecting ADR signals.ConclusionsImproved performance of ML models indicated that applying our model to EHR data is feasible and promising for detecting more accurate and comprehensive ADR signals.

BackgroundPrevious evidence suggests that there is a correlation between prolonged sitting time and cardio-metabolic disease, such as metabolic syndrome (MS). Cardiovascular disease is the second-leading cause of mortality in South Korea, a country with the longest working hours among all member states of the Organization for Economic Co-operation and Development. However, no previous study has investigated the relationships of overall sitting-time and occupation with MS in South Korea. Accordingly, the present study examined these relationships in a South Korean population.MethodsData from the sixth Korean National Health and Nutrition Examination Survey (KNHANES), a nationally representative survey with a cross-sectional design, were used in the present study. MS diagnoses were evaluated using the International Diabetes Foundation (IDF) criteria. Participants self-reported their overall sitting times, and occupations were classified using the Korean version of the Standard Classification of Occupations (KSCO). A multiple logistic regression analysis was conducted to evaluate the associations of sitting time and occupation with MS.ResultsThe risk of MS was 1.21-fold higher among participants who sat for >7 h/day than among those who sat for ≤7 h/day (odds ratio [OR]: 1.21, 95 % confidence interval [CI]: 1.00–1.46). Regarding occupation, office workers had a two-fold higher risk of MS than did agriculture, forestry, and fishery (AFF) workers (OR: 2.01, 95 % CI: 1.26–3.22). In a combined analysis of sitting time and occupation, male participants who sat for >7 h/day and reported an occupation that involves office work (OW) or machine fitting (MF) were significantly more likely to have MS when compared to those who sat for ≤7 h/day and were employed as AFF workers (>7 h/day × OW, OR: 2.41, 95 % CI: 1.05–5.51; >7 h/day × MF, OR: 2.92, 95 % CI: 1.43–5.93).ConclusionsExcessive sitting time and a sedentary occupation correlated positively with MS in South Korean adults. Accordingly, a reduction in the overall sitting time or inclusion of energy-expending activities in the workplace might improve the rate of MS.

IntroductionThe literature on stroke mortality and neighborhood effect is characterized by studies that are often Western society-oriented, with a lack of racial and cultural diversity. We estimated the effect of cross-level interaction between individual and regional socioeconomic status on the survival after onset of ischemic stroke.MethodsWe selected newly diagnosed ischemic stroke patients from 2002 to 2013 using stratified representative sampling data of 1,025,340 subjects. A total of 37,044 patients over the 10 years from 2004 to 2013 had newly diagnosed stroke. We calculated hazard ratios (HR) of 12- and 36-month mortality using the Cox proportional hazard model, with the reference group as stroke patients with high income in advantaged regions.ResultsFor the middle income level, the patients in advantaged regions showed low HRs for overall mortality (12-month HR 1.27; 95% confidence interval [CI], 1.13–1.44; 36-month HR 1.25; 95% CI, 1.14–1.37) compared to the others in disadvantaged regions (12-month HR 1.36; 95% CI, 1.19–1.56; 36-month HR 1.30; 95% CI, 1.17–1.44). Interestingly, for the low income level, the patients in advantaged regions showed high HRs for overall mortality (12-month HR 1.27; 95% CI, 1.13–1.44; 36-month HR 1.33; 95% CI, 1.22–1.46) compared to the others in disadvantaged regions (12-month HR 1.25; 95% CI, 1.09–1.43; 36-month HR 1.30; 95% CI, 1.18–1.44).ConclusionAlthough we need to perform further investigations to determine the exact mechanisms, regional deprivation, as well as medical factors, might be associated with survival after onset of ischemic stroke in low-income patients.

BackgroundThe aim of this study was to examine factors associated with the use of mental health consultation for depressive symptoms.MethodsWe used data from the 2013 Community Health Survey, which included responses from 13,269 individuals who reported that they had experienced depressive symptoms for more than 2 weeks in Korea. We investigated associations between mental health consultation rates for depressive symptoms and sociodemographic, socioeconomic, and health-related factors. Logistic regression analysis was used to examine the significance of associations.ResultsAmong participants who report depressive symptoms, 16.0% (n = 2120) undergo mental health consultation. Respondents with a college education or over are more likely to undergo mental health consultation (odds ratio (OR) = 1.49; 95% CI: 1.21–1.84) than respondents with less education. Individuals aged 70 years or above are less likely to receive mental health consultation than those aged between 19 and 29 years. Females exhibit higher mental health consultation rates than males. Respondents who are divorced show greater odds of receiving mental health consultation than respondents who are married and cohabitate with their spouse.ConclusionsThis study indicates that rates of use of mental health consultation services are lower among older adults and men and higher among divorced people. Educational level shows a significant positive association with mental health consultation among Koreans. The results could have implications for mental health policy in many ways in Korea.

Background: Sleep disturbance has been significantly associated with the incidence of cardiovascular disease (CVD) in the general population. However, despite the common prevalence of sleep disturbance in patients with type 2 diabetes, its relationship with the risk of CVD remains unclear. Here, we have examined the association of sleep disturbance with the incidence of all CVD and all-cause mortality in patients with newly-diagnosed type 2 diabetes. Methods: We used the Korean National Health Insurance Service-Health Screening Cohort data and included 36,058 patients with new-onset type 2 diabetes aged ≥ 40 years between 2004 and 2007, along with follow-up examinations to 2013. We used the ICD-10 code to measure sleep disturbance as a primary diagnosis and the multivariable Cox proportional hazards regression models to estimate the adjusted hazard ratio (AHR) and 95% confidence interval (CI) of all CVD, coronary heart disease (CHD), stroke, and all-cause mortality. Results: We identified 6897 cases of all CVD (CHD, n = 4138; stroke, n = 2759) and 2890 events of all-cause mortality during a mean follow-up period of 7.0 years. Sleep disturbance was associated with an increased risk of All CVD (AHR, 1.24; 95% CI, 1.06-1.46), CHD events (AHR, 1.24; 95% CI, 1.00-1.53), and all-cause mortality (AHR, 1.47; 95% CI, 1.15-1.87) in patients with new-onset type 2 diabetes. Furthermore, women (AHR, 1.33; 95% CI, 1.06-1.67) and middle-aged adults (AHR, 1.29; 95% CI, 1.02-1.64) with sleep disturbance had a significantly increased risk of CVD than those without; contrarily, men (AHR, 1.45; 95% CI, 1.09-1.95) and older adults (AHR, 1.51; 95% CI, 1.15-1.99) with sleep disturbance were associated with a significantly increased risk of all-cause mortality than those without. Conclusions: Our findings suggest that sleep disturbance is significantly associated with an increased risk of CVD and all-cause mortality in patients with new-onset type 2 diabetes.