Determinants of coronavirus disease 2019 infection by artificial intelligence technology: A study of 28 countries.

OBJECTIVES: The coronavirus disease 2019 pandemic has affected countries around the world since 2020, and an increasing number of people are being infected. The purpose of this research was to use big data and artificial intelligence technology to find key factors associated with the coronavirus disease 2019 infection. The results can be used as a reference for disease prevention in practice. METHODS: This study obtained data from the "Imperial College London YouGov Covid-19 Behaviour Tracker Open Data Hub", covering a total of 291,780 questionnaire results from 28 countries (April 1~August 31, 2020). Data included basic characteristics, lifestyle habits, disease history, and symptoms of each subject. Four types of machine learning classification models were used, including logistic regression, random forest, support vector machine, and artificial neural network, to build prediction modules. The performance of each module is presented as the area under the receiver operating characteristics curve. Then, this study further processed important factors selected by each module to obtain an overall ranking of determinants. RESULTS: This study found that the area under the receiver operating characteristics curve of the prediction modules established by the four machine learning methods were all >0.95, and the RF had the highest performance (area under the receiver operating characteristics curve is 0.988). Top ten factors associated with the coronavirus disease 2019 infection were identified in order of importance: whether the family had been tested, having no symptoms, loss of smell, loss of taste, a history of epilepsy, acquired immune deficiency syndrome, cystic fibrosis, sleeping alone, country, and the number of times leaving home in a day. CONCLUSIONS: This study used big data from 28 countries and artificial intelligence methods to determine the predictors of the coronavirus disease 2019 infection. The findings provide important insights for the coronavirus disease 2019 infection prevention strategies.

Atypical Teratoid/Rhabdoid Tumor in Taiwan: A Nationwide, Population-Based Study.

BACKGROUND: Atypical teratoid/rhabdoid tumor (AT/RT) is a rare, highly aggressive embryonal brain tumor most commonly presenting in young children. METHODS: We performed a nationwide, population-based study of AT/RT (ICD-O-3 code: 9508/3) in Taiwan using the Taiwan Cancer Registry Database and the National Death Certificate Database. RESULTS: A total of 47 cases (male/female = 29:18; median age at diagnosis, 23.3 months (IQR: 12.5-87.9)) were diagnosed with AT/RT between 1999 and 2014. AT/RT had higher prevalence in males (61.70%), in children < 36 months (55.32%), and at infratentorial or spinal locations (46.81%). Survival analyses demonstrated that patients ≥ 3 years of age (n = 21 (45%)) had a 5y-OS of 41% (p < 0.0001), treatment with radiotherapy only (n = 5 (11%)) led to a 5y-OS of 60%, treatment with chemotherapy with or without radiotherapy (n = 27 (62%)) was associated with a 5y-OS of 45% (p < 0.0001), and patients with a supratentorial tumor (n = 11 (23%)) had a 5y-OS of 51.95%. Predictors of better survival on univariate Cox proportional hazard modeling and confirmed with multivariate analysis included older age (≥1 year), supratentorial sites, and the administration of radiotherapy, chemotherapy, or both. Gender had no effect on survival. CONCLUSION: Older age, supratentorial site, and treatment with radiotherapy, chemotherapy, or both significantly improves the survival of patients with AT/RT.

Comparison of Acute and Chronic Surgical Complications Following Robot-Assisted, Laparoscopic, and Traditional Open Radical Prostatectomy Among Men in Taiwan.

Importance: Few studies have evaluated long-term surgical complications in patients with prostate cancer (PC) who receive open radical prostatectomy (ORP), laparoscopic radical prostatectomy (LRP), or robot-assisted radical prostatectomy (RARP). Objective: To examine the perioperative and postoperative surgical complications among patients with PC who underwent ORP, LRP, or RARP. Design, Setting, and Participants: This cohort study included patients who received a diagnosis of resectable PC and underwent RP between January 1 and December 31, 2015. Participants were enrolled in the Taiwan Cancer Registry. The index date was the date of surgery, and the follow-up duration was the period from the index date to December 31, 2018. Data analysis was performed in September 2020. Exposures: ORP, LRP, or RARP. Main Outcomes and Measures: Two multivariate mixed models accounting for hospital clusters were fitted to ascertain the association of RARP with treatment outcomes (ie, hospital stay, blood transfusion, postoperative pain, erectile dysfunction, urinary incontinence, and hernia); general linear regression models were used for continuous outcomes, the amount of blood transfused, and hospital stay, and logistic regression models were used for analyzing postoperative outcomes and surgical complications. Results: Of the 1407 patients included in this study, 315 (22.4%) received ORP (mean [SD] age, 66.4 [6.8] years), 276 (19.6%) received LRP (mean [SD] age, 66.8 [6.4] years), and 816 (58.0%) received RARP (mean [SD] age, 66.1 [6.7] years). Mean (SD) follow-up in the full cohort was 36.7 (4.6) months. No statistically significant differences were observed in age, clinical tumor stage, pathological tumor stage, Gleason score, Gleason grade group, preoperative prostate-specific antigen concentration, D'Amico risk classification, and hospital level. A shorter hospital stay was observed for patients undergoing RARP vs those undergoing ORP (mean [SE] difference, -1.64 [0.22] days; P < .001) and LRP (mean [SE] difference, -0.57 [0.23] days; P = .01). Patients undergoing RARP had lower odds of receiving a blood transfusion (RARP vs ORP: adjusted odds ratio [aOR], 0.25; 95% CI, 0.17-0.36; RARP vs LRP: aOR, 0.58; 95% CI, 0.37-0.91). For postoperative pain, RARP was associated with a decrease in the odds of moderate to severe postoperative pain for as long as 12 weeks compared with both ORP and LRP (eg, RARP vs LRP at week 12: aOR, 0.40; 95% CI, 0.19-0.85; P = .02). The aORs for RARP vs those for ORP and LRP in the third year after RP were, for erectile dysfunction, 0.74 (95% CI, 0.45-0.92) and 0.60 (95% CI, 0.36-0.98), respectively; for urinary incontinence, 0.93 (95% CI, 0.65-0.99) and 0.60 (95% CI, 0.42-0.86), respectively; and for hernia, 0.51 (95% CI, 0.31-0.84) and 0.82 (95% CI, 0.46-0.92), respectively. Conclusions and Relevance: In this study, undergoing RARP was associated with fewer acute and chronic postoperative complications than undergoing ORP or LRP.

Latest Comprehensive Medical Resource Consumption in Robot-Assisted versus Laparoscopic and Traditional Open Radical Prostatectomy: A Nationwide Population-Based Cohort Study.

BACKGROUND: Few studies have evaluated long-term medical monetary cost in patients with prostate cancer (PC) receiving open radical prostatectomy (ORP), laparoscopic radical prostatectomy (LRP), or robot-assisted radical prostatectomy (RARP). To the best of our knowledge, this is the largest and longest follow-up study to examine medical monetary cost in patients with PC undergoing ORP, LRP, or RARP. After adjustment for confounders, the medical monetary cost in the RARP group was the least compared with that in the ORP and LRP groups. PURPOSE: To estimate long-term medical resource consumption among patients with prostate cancer (PC) receiving open radical prostatectomy (ORP), laparoscopic radical prostatectomy (LRP), or robot-assisted radical prostatectomy (RARP). PATIENTS AND METHODS: Participants were men enrolled in the Taiwan Cancer Registry with localized PC diagnosis who received radical prostatectomy. After adjustment for confounders, a generalized linear mixed model was used to determine significant differences in the number of urology outpatient clinic visits required, proportion of patients being hospitalized for urinary diseases or surgical complications, and medical reimbursement for urinary diseases or surgical complications following ORP, LRP, or RARP in the first, second, and third years. Results: No differences were observed in the median number of urology outpatient clinic visits between the three types of surgical modalities up to the second year after ORP, LRP, and RARP (median: 15, 10, and seven visits, respectively; p < 0.001), but significant differences were observed in the third year. Similarly, with RARP (10.9% versus 18.7% in ORP and 9.8% in LRP; p = 0.0014), the rate of hospitalization for urinary diseases or surgical complications decreased in the third year. Medical reimbursement for urinary diseases or surgical complications reduced after RARP compared with that for ORP and LRP, with approximately 22% reduction in the first year (p = 0.0052) and 20-40% reduction in the third year (p value = 0.0024). CONCLUSIONS: Medical resource consumption in the RARP group was less compared with those in the ORP and LRP groups.

Oncologic Outcomes of Radical Prostatectomy and High-Dose Intensity-Modulated Radiotherapy with Androgen-Deprivation Therapy for Relatively Young Patients with Unfavorable Intermediate-Risk Prostate Adenocarcinoma.

PURPOSE: To estimate the oncologic outcomes of radical prostatectomy (RP) and high-dose intensity-modulated radiotherapy (IMRT) with short-term androgen-deprivation therapy (ADT) in relatively young men with unfavorable intermediate-risk prostate cancer, as defined by the National Comprehensive Cancer Network (NCCN-UIR-PC). PATIENTS AND METHODS: We enrolled relatively young men (≤65 years) from the Taiwan Cancer Registry who had been diagnosed as having NCCN-UIR-PC and who had received RP or high-dose IMRT (at least ≥72 Gy) with short-term ADT (4-6 months). After propensity score matching of the confounders, Cox proportional regression was used to model the time from the index date (i.e., date of diagnosis) to all-cause death, biochemical failure (BF), locoregional recurrence (LRR), and distant metastasis (DM). RESULTS: The corresponding adjusted hazard ratios (95% confidence intervals) of the risk of all-cause death, BF, LRR, and DM were 2.76 (1.36-5.60, p = 0.0050), 2.74 (1.72-4.84, p < 0.0001), 1.28 (1.09-1.90, p = 0.0324), and 2.11 (1.40-4.88, p = 0.0052), respectively. CONCLUSIONS: RP is superior to high-dose IMRT with short-term ADT in terms of oncologic outcomes for relatively young men with UIR-PC.

Longitudinal transition trajectory of gouty arthritis and its comorbidities: a population-based study.

The aim of the study was to investigate the longitudinal transition trajectory of gout and its comorbidities in male patients with gout in different age groups. A total of 3973 male patients who received a new diagnosis of gouty arthritis were identified from the Taiwan Longitudinal Health Insurance Database and divided into two age cohorts (<50 and ≥50 years). Each patient was individually followed from 2000 to 2009 to identify associated comorbidities, namely hypertension, hypercholesterolemia, diabetes mellitus, cardiovascular diseases, and chronic kidney disease. Two outcome measurements of stroke and all-cause cancer were further identified until 2010. The transition trajectory was divided into the following five phenotype groups: persistent hypertension combined with a high prevalence of various gout-related comorbidities, persistent hypercholesterolemia combined with a moderate prevalence of various gout-related comorbidities, persistent low prevalence of various gout-related comorbidities, moderate to high prevalence of various gout-related comorbidities, and low to high prevalence of various gout-related comorbidities. Although the younger and older patients had a similar longitudinal transition trajectory of gout-related comorbidities, the older patients had a higher 10-year likelihood of transition from a low or moderate to a high prevalence of various gout-related comorbidities. In addition, the incidences of stroke and all-cause cancer were higher in the groups with high and moderate to high prevalences of various gout-related comorbidities than in the other groups. The occurrence of gouty arthritis in different life stages can cause cluster effects involving varying degrees of comorbidities over time. The findings of the current study can provide additional knowledge and increase clinical awareness regarding the early assessment and management of gout-related comorbidities in clinical practice.

Cancer risk in HBV patients with statin and metformin use: a population-based cohort study.

Chronic infection with hepatitis B virus (HBV) often causes chronic inflammation of the liver with an increased incidence of hepatocellular carcinoma (HCC). HBV-infected individuals may also have an increased incidence of nonliver cancers. Taking statin or metformin may decrease inflammation and infiltration, which may, as a result, reduce the risk of liver cancer or other major cancers in patients with HBV infection. The purpose of this study was to evaluate the hypothesis that statin and metformin could reduce the incidence of liver cancer (HCC) or nonliver cancers in patients with HBV.Using the Taiwan Longitudinal Health Insurance Database 2000 to 2008, this cohort study comprised patients with a recorded diagnosis of HBV (N = 71,847) between January 1, 2000 and December 31, 2008. Each patient was followed-up until the end of 2008. The occurrence of HCC or a nonliver cancer was evaluated in patients who either were or were not taking statin or metformin. Cox proportional hazard regressions were used to evaluate the cancer incidence after adjusting for known confounding factors.In total, 71,824 HBV-infected patients comprised the study cohort. Our study showed that either metformin or statin use was associated with a reduction in the incidence of cancer. This was most prominent in patients taking both statin and metformin. The adjusted hazard ratios (HRs) for patients using only statin were 0.52 (95% confidence interval [CI], 0.48-0.57) for all cancers, 0.28 (95% CI, 0.23-0.35) for liver cancer, and 0.63 (95% CI, 0.57-0.70) for nonliver cancers. Patients taking only metformin had risk-adjusted HRs of 0.82 (95% CI, 0.75-0.90) for all cancers, 0.97 (95% CI, 0.84-1.14) for liver cancer, and 0.75 (95% CI, 0.67-0.84) for nonliver cancers. A dose-dependent effect of statin use for chemoprevention was observed for all cancers, including both liver cancer and nonliver cancers. A dose-dependent effect of metformin was also seen in liver cancer and nonliver cancers without stratification into different cumulative daily doses of statin use.This population-based cohort study investigated the protective effect of statin and metformin against cancer events in patients with HBV infection. Our study demonstrated that either statin or metformin served as independent chemopreventive agents with a dose-response effect in reducing the incidence of cancer with a dose-response effect of the agents and an additive or synergistic effect of combining statin and metformin use in reducing the incidence of many cancers.

Comorbidity as an independent risk factor in patients with cancer: an 8-year population-based study.

This study determined the prevalence of medical conditions in patients with cancer and their impact on outcome. We evaluated a cohort of 37,411 patients diagnosed with cancer between 2000 and 2008 in Taiwan, collecting the cancer diagnosis and chronic disease diagnoses. The severity of the comorbid condition was correlated with the cancer diagnosis and outcome. Overall, 71.9% of the study population had 1 or more comorbid conditions. Patients with none (n = 10 508), 1 (n = 8881), 2 (n = 6583), and 3 or more (n = 11 439) comorbid conditions had mortality rates of 11.49%, 15.99%, 19.61%, and 29.39%, respectively. Older patients with comorbid conditions had a significantly higher chance of death. Dementia, heart disease, or cerebrovascular diseases were associated with the highest mortality. Cancer patients with comorbid conditions have a significantly higher risk of death. Prevention and better medical management of comorbid conditions is likely to result in improved outcomes for patients with cancer.

Emergency department utilization can indicate early diagnosis of digestive tract cancers: A population-based study in Taiwan.

BACKGROUND AND OBJECTIVE: Patients who visit emergency department (ED) may have symptoms of occult cancers. METHODS: We studied a random cohort of one million subjects from Taiwan National Health Insurance Research Database between 2000 and 2008 to evaluate the ED utilization of individuals who were subsequently diagnosed with digestive tract cancers. The case group was digestive tract cancer patients and the control group was traumatic fracture patients. We reviewed record of ED visits only from 4 to 15 months before the cancer diagnoses. RESULTS: There were 2635 and 6665 in the case and control groups respectively. Patients' adjusted odds ratio with 95% confidence interval for the case group were 1.36 (1.06-1.74) for Abdominal ultrasound, 2.16 (1.61-2.90) pan-endoscopy, 1.72 (1.33-2.22) guaiac fecal-occult blood test, 1.42 (1.28-1.58) plain abdominal X-rays, 1.20 (1.09-1.32) SGOT, 1.27 (1.14-1.40) SGPT, 1.66 (1.41-1.95) total bilirubin, 2.41 (1.89-3.08) direct bilirubin, 1.21 (1.01-1.46) hemoglobin and 3.63 (2.66-4.94) blood transfusion, respectively. Blood transfusion in the ED was a significant predictor of the individual subsequently diagnosed with digestive tract cancer. CONCLUSIONS: The health system could identify high risk patients early by real-time review of their ED utilization before the diagnosis of digestive tract cancers. We proposed a follow-up methodology for daily screening of patients with high risk of digestive tract cancer by informatics system in the ED.

Critical laboratory result reporting system in cancer patients.

HYPOTHESIS: Automatic transmission of computer-generated Critical Laboratory Result Reports (CLRRs) to physicians can improve the care of advanced cancer patients by improving the communication efficacy of important medical information. METHOD: We followed a cohort of 2012 cancer patients from diagnosis to five years or to death if it occurred before five years from diagnosis. The incidence and number of CLRRs and their association with diagnosis, age, gender, tumor size, and clinical staging were evaluated. The CLRRs that were reported included for example: glucose<40 or>500, hemoglobin<6.0. (Appendix 1) RESULTS: Two thousand, twelve patients with cancer were included in the study; 45.6 percent had one or more critical laboratory results that required a CLRR. Older patients greater than or equal to 75 years of age had more CLRRs than younger patients. Patients with colorectal, liver, and lung cancer had a significantly higher number of CLRRs. More CLRRs were also seen in late-staged cancers. These conditions also have higher mortality rates. CONCLUSION: Critical values are common in patients with cancer, especially older patients with advanced disease. They occur more commonly with some cancers of liver and lung cancers. Our data demonstrate that critical laboratory values can be transmitted successfully to physicians. The impact of this system promises to improve the care of these individuals' serious illnesses. A prospective study to demonstrate the benefit of this system is being planned.