ECG-surv: A deep learning-based model to predict time to 1-year mortality from 12-lead electrocardiogram.

BACKGROUND: Electrocardiogram (ECG) abnormalities have demonstrated potential as prognostic indicators of patient survival. However, the traditional statistical approach is constrained by structured data input, limiting its ability to fully leverage the predictive value of ECG data in prognostic modeling. METHODS: This study aims to introduce and evaluate a deep-learning model to simultaneously handle censored data and unstructured ECG data for survival analysis. We herein introduce a novel deep neural network called ECG-surv, which includes a feature extraction neural network and a time-to-event analysis neural network. The proposed model is specifically designed to predict the time to 1-year mortality by extracting and analyzing unique features from 12-lead ECG data. ECG-surv was evaluated using both an independent test set and an external set, which were collected using different ECG devices. RESULTS: The performance of ECG-surv surpassed that of the Cox proportional model, which included demographics and ECG waveform parameters, in predicting 1-year all-cause mortality, with a significantly higher concordance index (C-index) in ECG-surv than in the Cox model using both the independent test set (0.860 [95% CI: 0.859- 0.861] vs. 0.796 [95% CI: 0.791- 0.800]) and the external test set (0.813 [95% CI: 0.807- 0.814] vs. 0.764 [95% CI: 0.755- 0.770]). ECG-surv also demonstrated exceptional predictive ability for cardiovascular death (C-index of 0.891 [95% CI: 0.890- 0.893]), outperforming the Framingham risk Cox model (C-index of 0.734 [95% CI: 0.715-0.752]). CONCLUSION: ECG-surv effectively utilized unstructured ECG data in a survival analysis. It outperformed traditional statistical approaches in predicting 1-year all-cause mortality and cardiovascular death, which makes it a valuable tool for predicting patient survival.

Diabetic peripheral neuropathy: age-stratified glycemic control.

Background: We explore the effect of suboptimal glycemic control on the incidence of diabetic peripheral neuropathy (DPN) in both non-elderly and elderly patients with type 2 diabetes mellitus (T2DM). Methods: A 6-year follow-up study (2013-2019) enrolled T2DM patients aged >20 without DPN. Participants were classified into two groups: those below 65 years (non-elderly) and those 65 years or older (elderly). Biochemical measurements, including glycated hemoglobin (HbA1C), were recorded regularly. DPN was diagnosed using the Michigan Neuropathy Screening Instrument examination. The outcome was DPN occurrence in 2019. Results: In 552 enrollments (69% non-elderly), DPN occurred in 8.4% non-elderly and 24.0% elderly patients. A higher initial HbA1C level was significantly linked with a higher risk of future DPN in the non-elderly group (adjusted odds ratio [AOR] 1.46, 95% CI 1.13-1.89, p=0.004). In comparison, HbA1c at the end of the study period was not associated with DPN in the non-elderly group (AOR 1.17, 95% CI 0.72-1.90, p=0.526). In the elderly group, no statistical relationship was found between HbA1C levels and DPN, either in 2013 or in 2019. Conclusion: Suboptimal glycemic control at baseline, rather than at the end of the study period, predicts an increased risk of future DPN in individuals with T2DM under age 65. This correlation is not seen in elderly patients. Therefore, we recommend implementing enhanced glycemic control early in middle-aged T2DM patients and propose individualized therapeutic strategies for diabetes in different age groups.

An efficient deep neural network for automatic classification of acute intracranial hemorrhages in brain CT scans.

BACKGROUND: Recent advancements in deep learning models have demonstrated their potential in the field of medical imaging, achieving remarkable performance surpassing human capabilities in tasks such as classification and segmentation. However, these modern state-of-the-art network architectures often demand substantial computational resources, which limits their practical application in resource-constrained settings. This study aims to propose an efficient diagnostic deep learning model specifically designed for the classification of intracranial hemorrhage in brain CT scans. METHOD: Our proposed model utilizes a combination of depthwise separable convolutions and a multi-receptive field mechanism to achieve a trade-off between performance and computational efficiency. The model was trained on RSNA datasets and validated on CQ500 dataset and PhysioNet dataset. RESULT: Through a comprehensive comparison with state-of-the-art models, our model achieves an average AUROC score of 0.952 on RSNA datasets and exhibits robust generalization capabilities, comparable to SE-ResNeXt, across other open datasets. Furthermore, the parameter count of our model is just 3 % of that of MobileNet V3. CONCLUSION: This study presents a diagnostic deep-learning model that is optimized for classifying intracranial hemorrhages in brain CT scans. The efficient characteristics make our proposed model highly promising for broader applications in medical settings.

Polygenic Risk Score Predicts Earlier-Onset Adult Systemic Lupus Erythematosus and First-Year Renal Diseases in a Taiwanese Cohort.

OBJECTIVES: This study aimed to develop a predictive model using polygenic risk score (PRS) to forecast renal outcomes for adult systemic lupus erythematosus (SLE) in a Taiwanese population. METHODS: Patients with SLE (n=2782) and matched non-SLE controls (n=11 128) were genotyped using Genome-Wide TWB 2.0 single-nucleotide polymorphism (SNP) array. PRS models (C+T, LDpred2, Lassosum, PRSice-2, PRS-continuous shrinkage (CS)) were constructed for predicting SLE susceptibility. Logistic regression was assessed for C+T-based PRS association with renal involvement in patients with SLE. RESULTS: In the training set, C+T-based SLE-PRS, only incorporating 27 SNPs, outperformed other models with area under the curve (AUC) values of 0.629, surpassing Lassosum (AUC=0.621), PRSice-2 (AUC=0.615), LDpred2 (AUC=0.609) and PRS-CS (AUC=0.602). Additionally, C+T-based SLE-PRS demonstrated consistent predictive capacity in the testing set (AUC=0.620). Individuals in the highest quartile exhibited earlier SLE onset (39.06 vs 44.22 years, p<0.01), higher Systemic Lupus Erythematosus Disease Activity Index scores (3.00 vs 2.37, p=0.04), elevated risks of renal involvement within the first year of SLE diagnosis, including WHO class III-IV lupus nephritis (OR 2.36, 95% CI 1.47 to 3.80, p<0.01), estimated glomerular filtration rate <60 mL/min/1.73m2 (OR 1.49, 95% CI 1.18 to 1.89, p<0.01) and urine protein-to-creatinine ratio >150 mg/day (OR 2.07, 95% CI 1.49 to 2.89, p<0.01), along with increased seropositivity risks, compared with those in the lowest quartile. Furthermore, among patients with SLE with onset before 50 years, the highest PRS quartile was significantly associated with more serious renal diseases within the first year of SLE diagnosis. CONCLUSIONS: PRS of SLE is associated with earlier onset, renal involvement within the first year of SLE diagnosis and seropositivity in Taiwanese patients. Integrating PRS with clinical decision-making may enhance lupus nephritis screening and early treatment to improve renal outcomes in patients with SLE.

Predicting ischemic stroke patients' prognosis changes using machine learning in a nationwide stroke registry.

Accurately predicting the prognosis of ischemic stroke patients after discharge is crucial for physicians to plan for long-term health care. Although previous studies have demonstrated that machine learning (ML) shows reasonably accurate stroke outcome predictions with limited datasets, to identify specific clinical features associated with prognosis changes after stroke that could aid physicians and patients in devising improved recovery care plans have been challenging. This study aimed to overcome these gaps by utilizing a large national stroke registry database to assess various prediction models that estimate how patients' prognosis changes over time with associated clinical factors. To properly evaluate the best predictive approaches currently available and avoid prejudice, this study employed three different prognosis prediction models including a statistical logistic regression model, commonly used clinical-based scores, and a latest high-performance ML-based XGBoost model. The study revealed that the XGBoost model outperformed other two traditional models, achieving an AUROC of 0.929 in predicting the prognosis changes of stroke patients followed for 3 months. In addition, the XGBoost model maintained remarkably high precision even when using only selected 20 most relevant clinical features compared to full clinical datasets used in the study. These selected features closely correlated with significant changes in clinical outcomes for stroke patients and showed to be effective for predicting prognosis changes after discharge, allowing physicians to make optimal decisions regarding their patients' recovery.

Antibiotic exposure during pregnancy increases risk for childhood atopic diseases: a nationwide cohort study.

PURPOSE: The prevalence of atopic diseases has increased in recent decades. A possible link between antibiotic use during pregnancy and childhood atopic disease has been proposed. The aim of this study is to explore the association of antibiotic exposure during pregnancy with childhood atopic diseases from a nationwide, population-based perspective. METHODS: This was a nationwide population-based cohort study. Taiwan's National Health Insurance Research Database was the main source of data. The pairing of mothers and children was achieved by linking the NHIRD with the Taiwan Maternal and Child Health Database. This study enrolled the first-time pregnancies from 2004 to 2010. Infants of multiple delivery, preterm delivery, and death before 5 years old were excluded. All participants were followed up at least for 5 years. Antenatal antibiotics prescribed to mothers during the pregnancy period were reviewed. Children with more than two outpatient visits, or one admission, with a main diagnosis of asthma, allergic rhinitis, or atopic dermatitis were regarded as having an atopic disease. RESULTS: A total of 900,584 children were enrolled in this study. The adjusted hazard ratios of antibiotic exposure during pregnancy to childhood atopic diseases were 1.12 for atopic dermatitis, 1.06 for asthma, and 1.08 for allergic rhinitis, all of which reached statistical significance. The trimester effect was not significant. There was a trend showing the higher the number of times a child was prenatally exposed to antibiotics, the higher the hazard ratio was for childhood atopic diseases. CONCLUSIONS: Prenatal antibiotic exposure might increase the risk of childhood atopic diseases in a dose-dependent manner.

The use of individual-based FDG-PET volume of interest in predicting conversion from mild cognitive impairment to dementia.

BACKGROUND: Based on a longitudinal cohort design, the aim of this study was to investigate whether individual-based 18F fluorodeoxyglucose positron emission tomography (18F-FDG-PET) regional signals can predict dementia conversion in patients with mild cognitive impairment (MCI). METHODS: We included 44 MCI converters (MCI-C), 38 non-converters (MCI-NC), 42 patients with Alzheimer's disease with dementia, and 40 cognitively normal controls. Data from annual cognitive measurements, 3D T1 magnetic resonance imaging (MRI) scans, and 18F-FDG-PET scans were used for outcome analysis. An individual-based FDG-PET approach was applied using seven volumes of interest (VOIs), Z transformed using a normal FDG-PET template. Hypometabolism was defined as a Z score < -2 of regional standard uptake value ratio. For the longitudinal cognitive test scores, generalized estimating equations were used. A linear mixed-effects model was used to compare the temporal impact of cortical hypometabolism and cortical thickness degeneration. RESULTS: The clinical follow-up period was 6.6 ± 3.8 years (range 3.1 to 16.0 years). The trend of cognitive decline could differentiate MCI-C from MCI-NC after 3 years of follow-up. In the baseline 18F-FDG-PET scan of the patients with MCI, medial temporal lobe (MTL; 94.7% sensitivity, 80.5% specificity) and posterior cingulate cortex (PCC; 89.5% sensitivity, 73.1% specificity) hypometabolism predicted conversion with high accuracy. 18F-FDG-PET hypometabolism preceded dementia conversion at an interval of 3.70 ± 1.68 years and was earlier than volumetric changes, with the exception of the MTL. CONCLUSIONS: Our finding supports the use of individual-based 18F-FDG-PET analysis to predict MCI conversion to dementia. Reduced FDG-PET metabolism in the MTL and PCC were strongly associated with future cognitive decline in the MCI-C group. Changes in 18F-FDG-PET occurred 1 to 8 years prior to conversion to dementia. Progressive hypometabolism in the PCC, precuneus and lateral temporal lobe, but not MTL, preceded MRI findings at the MCI stage.

Machine learning models for predicting unscheduled return visits of patients with abdominal pain at emergency department and validation during COVID-19 pandemic: A retrospective cohort study.

Machine learning (ML) models for predicting 72-hour unscheduled return visits (URVs) for patients with abdominal pain in the emergency department (ED) were developed in a previous study. This study refined the data to adjust previous prediction models and evaluated the model performance in future data validation during the COVID-19 era. We aimed to evaluate the practicality of the ML models and compare the URVs before and during the COVID-19 pandemic. We used electronic health records from Chang Gung Memorial Hospital from 2018 to 2019 as a training dataset, and various machine learning models, including logistic regression (LR), random forest (RF), extreme gradient boosting (XGB), and voting classifier (VC) were developed and subsequently used to validate against the 2020 to 2021 data. The models highlighted several determinants for 72-hour URVs, including patient age, prior ER visits, specific vital signs, and medical interventions. The LR, XGB, and VC models exhibited the same AUC of 0.71 in the testing set, whereas the VC model displayed a higher F1 score (0.21). The XGB model demonstrated the highest specificity (0.99) and precision (0.64) but the lowest sensitivity (0.01). Among these models, the VC model showed the most favorable, balanced, and comprehensive performance. Despite the promising results, the study illuminated challenges in predictive modeling, such as the unforeseen influences of global events, such as the COVID-19 pandemic. These findings not only highlight the significant potential of machine learning in augmenting emergency care but also underline the importance of iterative refinement in response to changing real-world conditions.

The impact of sodium-glucose co-transporter-2 inhibitors on dementia and cardiovascular events in diabetic patients with atrial fibrillation.

AIMS: The effectiveness of sodium-glucose co-transporter-2 inhibitors (SGLT2i) on incident dementia in patients with diabetes and atrial fibrillation (AF) remains unknown. This study aimed to investigate the association between SGLT2i and the risk of incident dementia in diabetic patients with AF, and to explore the interactions with oral anticoagulants or dipeptidyl peptidase-4 inhibitors (DPP4i). MATERIALS AND METHODS: We conducted a cohort study using Taiwan's National Health Insurance Research Database. Patients with diabetes and AFwithout a prior history of established cardiovascular diseases, were identified. Using propensity score matching, 810 patients receiving SGLT2i were matched with 1620 patients not receiving SGLT2i. The primary outcome was incident dementia, and secondary outcomes included composite cardiovascular events and mortality. RESULTS: After up to 5 years of follow-up, SGLT2i use was associated with a significantly lower risk of incident dementia (hazard: 0.71, 95% confidence interval: 0.51-0.98), particularly vascular dementia (HR: 0.44, 95% CI: 0.24-0.82). SGLT2i was related to reduced risks of AF-related hospitalisation (HR: 0.72, 95% CI: 0.56-0.93), stroke (HR: 0.75, 95% CI: 0.60-0.94), and all-cause death (HR: 0.33, 95% CI: 0.24-0.44). The protective effects were consistent irrespective of the concurrent use of non-vitamin K antagonist oral anticoagulants (NOACs) or DPP4i. CONCLUSIONS: In diabetic patients with AF, SGLT2i was associated with reduced risks of incident dementia, AF-related hospitalisation, stroke, and all-cause death. The protective effects were independent of either concurrent use of NOACs or DPP4i.

Increased NAFLD risk in newly diagnosed patients with RA during the first 4 years of follow-up: a nationwide, population-based cohort study.

BACKGROUND: Although the non-alcoholic fatty liver disease (NAFLD) is prevalent in the general population, NAFLD risk in newly diagnosed rheumatoid arthritis (RA) has rarely been explored. In this population-based cohort, we examined NAFLD risk in patients with RA and identified the potential risk factors. DESIGN: Retrospective study. SETTING: Taiwan. PARTICIPANTS: 2281 newly diagnosed patients with RA and selected 91 240 individuals without RA to match with patients with RA (1:40) by age, gender, income status and urbanisation level of the residence. OUTCOMES: In this retrospective study using the 2000-2018 claim data from two-million representative Taiwanese population, we identified and compared the incidence rates (IRs) of NAFLD and alcoholic fatty liver disease (AFLD) between RA and non-RA groups. Using multivariable regression analyses, we estimated adjusted HR (aHR) of NAFLD development in patients with RA compared with individuals without RA, with 95% CIs. RESULTS: The incidences of NALFD and AFLD were not significantly different between individuals with RA and without RA during the 17-year follow-up period. However, patients with RA had significantly increased NAFLD risk during the first 4 years after RA diagnosis, with IR ratio of 1.66 fold (95% CI 1.18 to 2.33, p<0.005), but the risk was reduced after the first 4 years. Multivariable regression analyses revealed that aHR was 2.77-fold greater in patients not receiving disease-modifying anti-rheumatic drugs therapy than in non-RA subjects (p<0.05). Old age, women, low-income status and obesity could significantly predict NAFLD development. CONCLUSIONS: We demonstrated elevated risk of NAFLD in patients with RA during the first 4 years after RA diagnosis, and old age, women, low-income status and obesity were significant predictors of NAFLD.

Different dimensions of daytime sleepiness predicted mortality in older adults: Sex and muscle power-specific risk in Yilan Study, Taiwan.

OBJECTIVES: This study aimed to investigate the relationship between daytime sleepiness and mortality risk among older adults. The moderating effects of sex and physical function were examined. METHODS: This 9-year follow-up study was conducted with community-dwelling individuals aged ≥65 years. Daytime sleepiness was evaluated using the Epworth Sleepiness Scale (ESS). Exploratory factor analysis (EFA) was used to examine the ESS factors. Handgrip strength was measured to assess physical function, and the highest quartile was defined as good muscle power. Cox regression analysis was used to estimate the 9-year all-cause mortality risk. The interaction terms were examined to evaluate their moderating effect. RESULTS: In total, 2588 individuals participated in the study. The EFA explored two factors: the passive factor (PF) and the active factor (AF). After controlling for various covariates, the cutoff-defined daytime sleepiness (ESS≥11), total raw scores, and factor scores of the ESS all failed to predict mortality risk. The 3-way interaction terms showed statistical significance in terms of [sex × PF × muscle power (p = 0.03)] but not for [sex × AF × muscle power (p = 0.11)]. Specifically, PF predicted mortality risk in women with good muscle power (hazard ratio (HR): 1.48; 95 % confidence interval (CI): 1.04-2.10), which is female-specific. In contrast, AF predicted mortality risk only in men with good muscle power (HR: 1.35; 95 % CI: 1.02-1.78). CONCLUSIONS: The ESS-measured daytime sleepiness in older adults is multidimensional. The mortality risk for each dimension was determined based on sex and physical function.

Factors associated with mental illness in patients with rheumatoid arthritis initiating b/ts DMARDs: A population-based study.

AIM: Mental health is an essential issue in patients with rheumatoid arthritis (RA) but remains unclear among those receiving biological and targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs). We aim to assess the incidence and factors associated with mental illness among patients with RA who underwent b/tsDMARD therapy. METHOD: We used Taiwan's National Health Insurance Research Database for the period 2001-2020 to identify patients with RA receiving b/tsDMARDs. The primary outcome was newly developed mental illness, including anxiety and mood disorders. We performed a Cox regression analysis to determine factors associated with mental illness and presented as hazard ratios (HR) with 95% confidence interval (CI). RESULTS: We enrolled 10 852 patients, with 7854 patients receiving tumor necrosis factors inhibitors (TNFi), 1693 patients receiving non-TNFi bDMARDs, and 1305 patients treated with tsDMARD. We found that 13.62% of enrolled patients developed mental illness, with an incidence rate of 4054 per 100 000 person-year. Those receiving tocilizumab (aHR 0.64, 95% CI: 0.51-0.82), abatacept (aHR 0.69, 95% CI: 0.55-0.86), or tsDMARDs (aHR 0.58, 95% CI: 0.47-0.73) had a lower risk of mental illness compared with those receiving TNFi. We also found that old age, low income, diabetes mellitus, use of cyclosporine, and use of steroids were associated with incident mental illness. CONCLUSION: This population-based study investigated the incidence and factors associated with mental illness among patients with RA receiving b/tsDMARDs. Our findings highlight the need for vigilance with respect to the possibility of mental illness in patients with RA.

Association between HNF4A rs1800961 polymorphisms and gallstones in a Taiwanese population.

BACKGROUND AND AIM: A large genetic effect of a novel gallstone-associated genetic variant, the hepatocyte nuclear factor 4α (HNF4A) rs1800961 polymorphism, has been identified through recent genome-wide association studies. However, this effect has not been validated in Asian populations. We investigated the association between the rs1800961 variant and gallstones among a Taiwanese population. METHODS: A total of 20 405 participants aged between 30 and 70 years voluntarily enrolled in the Taiwan Biobank. Self-report questionnaires, physical examinations, biochemical tests, and genotyping were used for analysis. The association of the HNF4A rs1800961 variant and other metabolic risks with gallstone disease was analyzed using multiple logistic regression models. RESULTS: The minor T allele of HNF4A rs1800961 was associated with an increased risk of gallstone, and the association remained significant even after adjustment for other risk factors including age, body mass index (BMI), diabetes, hyperlipidemia, hypertension, and cigarette smoking (adjusted odds ratio [OR] = 1.90, 95% confidence interval [CI] = 1.31 to 2.75) in male participants. When further stratified by BMI and age, the lithogenic effect was the most significant in male participants with obesity (adjusted OR = 3.55, 95% CI = 1.92 to 6.56) and who were younger (adjusted OR = 2.45, 95% CI = 1.49 to 4.04). CONCLUSION: The novel gallstone-associated HNF4A rs1800961 variant was associated with the risk of gallstone in the Taiwanese men. Screening for the rs1800961 polymorphism may be particularly useful in assessing the risk of gallstone formation in younger or obese men.

Ten-year follow-up investigation of stroke risk in systemic lupus erythematosus.

BACKGROUND AND PURPOSE: To analyse the long-term risk of ischaemic stroke and the clinical effects of antithrombotics on the risk of haemorrhagic stroke in patients with systemic lupus erythematosus (SLE). METHODS: A retrospective cohort study was conducted using a population-based database taken from Taiwan National Health Insurance Research Database. Patients with SLE between 2000 and 2008 were registered and matched with two controls by the index date, age, gender and Charlson Comorbidity Index (CCI). These subjects were followed until either stroke event or 31 December 2013. Adjusted HRs (aHRs) for strokes were estimated with Cox regression models, and the cumulative incidence of ischaemic stroke was analysed by log-rank test and Kaplan-Meier survival analysis. RESULTS: In total, 8310 patients with SLE and 16 620 patients without SLE were included. In general, patients with SLE had higher rates of ischaemic stroke (5.4% vs 3.3%) and haemorrhagic stroke (1.5% vs 0.6%) than in controls. In multivariate analysis adjusted to age, gender, CCI, urbanisation level and antithrombotics uses, aHRs of all strokes, ischaemic stroke and haemorrhagic stroke were 1.73 (95% CI: 1.54 to 1.94), 1.65 (95% CI: 1.45 to 1.87) and 2.24 (95% CI: 1.71 to 2.95), respectively, in patients with SLE. Patients with SLE were significantly more likely to suffer ischaemic stroke than patients without SLE, even 10 years after SLE diagnosis (6.12% vs 3.50%, p<0.001). Antiplatelet use increased the risk of haemorrhagic stroke in SLE group (aHR=1.74, 95% CI: 1.18 to 2.57). CONCLUSIONS: Patients with SLE are at greater risk of developing ischaemic stroke that lasts for 10 years. Antiplatelets should be carefully administered to prevent cardiovascular events in patients with SLE due to the risk of haemorrhagic stroke.

Maternal risk factors for preterm birth in Taiwan, a nationwide population-based cohort study.

BACKGROUND: The rate of preterm birth is increasing globally. It causes significant short-term and long-term health care burdens. A comprehensive recognition of the risk factors related to preterm births is important in the prevention of preterm birth. Our study is to investigate the incidence and maternal risk factors of preterm birth from a nationwide population-based perspective. METHODS: This is a retrospective cohort study. All live births from 2004 to 2014 in Taiwan enrolled. The main data source was Taiwan's Birth Certificate Application (BCA) database. The BCA database was linked with the National Health Insurance Research Database (NHIRD) to establish any links between information on newborns and maternal underlying disease. RESULTS: A total of 1,385,979 births were included in the analysis. The incidence of preterm birth increased gradually in Taiwan from 8.85% in 2004 to 10.73% in 2014. Maternal age, socioeconomic status, maternal allergy and autoimmune diseases, gynecological diseases, and pregnancy-related complications were significant risk factors for preterm birth. CONCLUSION: The overall incidence of preterm births has gradually increased in Taiwan. Maternal age, socioeconomic status, certain underlying diseases, and pregnancy-related complications were risk factors for preterm birth.

Development of a Machine Learning Model to Predict the Use of Surgery in Patients With Rheumatoid Arthritis.

OBJECTIVE: One in five patients with rheumatoid arthritis (RA) rely on surgery to restore joint function. However, variable response to disease-modifying antirheumatic drugs (DMARDs) complicates surgical planning, and it is difficult to predict which patients may ultimately require surgery. We used machine learning to develop predictive models for the likelihood of undergoing an operation related to RA and which type of operation patients who require surgery undergo. METHODS: We used electronic health record data to train two extreme gradient boosting machine learning models. The first model predicted patients' probabilities of undergoing surgery ≥5 years after their initial clinic visit. The second model predicted whether patients who underwent surgery would undergo a major joint replacement versus a less intensive procedure. Predictors included demographics, comorbidities, and medication data. The primary outcome was model discrimination, measured by area under the receiver operating characteristic curve (AUC). RESULTS: We identified 5,481 patients, of whom 278 (5.1%) underwent surgery. There was no significant difference in the frequency of DMARD or steroid prescriptions between patients who did and did not have surgery, though nonsteroidal anti-inflammatory drug prescriptions were more common among patients who did have surgery (P = 0.03). The model predicting use of surgery had an AUC of 0.90 ± 0.02. The model predicting type of surgery had an AUC of 0.58 ± 0.10. CONCLUSIONS: Predictive models using clinical data have the potential to facilitate identification of patients who may undergo rheumatoid-related surgery, but not what type of procedure they will need. Integrating similar models into practice has the potential to improve surgical planning.

Gestational hypertension and risk of atopic diseases in offspring, a national-wide cohort study.

Introduction: Gestational hypertension can lead to complications, such as preeclampsia. Preeclampsia is one of leading causes of perinatal morbidity and mortality. Abnormal placentation, immune dysregulation, and maternal inflammatory response are possible etiologies. The burden of atopic diseases is increasing worldwide. Prenatal exposure might play a role in the pathogenesis of these two diseases. The aim of this study was to evaluate the association between gestational hypertension and atopic diseases from a nationwide perspective. Material and methods: The primary data were retrieved from Taiwan's National Health Insurance Research Database. The Maternal and Child Health Database was used to generate links between mothers and children. From 2004 to 2019, mothers with a diagnosis of gestational hypertension were identified as cases. The control groups were matched to the cases by maternal age, neonatal gender, date of birth, at a control-to-case ratio of 4:1. Each child was reviewed to confirm the diagnosis of atopic disease. Covariates including both maternal and neonatal factors were also collected. Results: A total of 1,935,874 primiparas were enrolled in this study. After excluding 16,851 mothers with a history of hypertension, a total of 1,919,023 offspring were included in the study for the period 2004-2019. Gestational hypertension was associated with asthma (HR, 1.12, 95% CI, 1.02-1.23) and atopic dermatitis (HR, 1.10, 95% CI, 1.00-1.21) in offspring after controlling for cofactors. Nevertheless, gestational hypertension did not play an independent factor for allergic rhinitis (HR, 1.02, 95% CI, 0.95-1.10) or urticaria (HR, 1.02, 95% CI, 0.91-1.15). Conclusion: Maternal gestational hypertension increases the cumulative risk for asthma and atopic dermatitis in offspring.

Association between periodontitis treatment and dementia in Taiwanese adults.

BACKGROUND: The chronic systemic inflammatory response in periodontitis may be a potential risk factor for dementia, especially in adults. This study determined the association between periodontal treatment and dementia in adults and evaluated the effect of regular scaling treatment on the risk of dementia in this population. METHODS: This case-control study identified 18,930 patients with a dementia-related diagnosis from the Taiwan National Health Insurance Research Database. Scaling and periodontal emergency treatments were evaluated after 1 year and 3 years. Using multivariable logistic regression analysis to evaluate the association between periodontal emergency treatment and dementia risk. RESULTS: The results showed that scaling treatment rates were lower in the dementia cohort than the non-dementia cohort after 1 and 3 years. Patients who received periodontal emergency treatment within 3 years had a significantly increased risk of dementia. Furthermore, patients with periodontitis who did not receive scaling treatment within 3 years had a higher risk of dementia than patients without periodontitis (OR, 1.22; 95% CI, 1.10-1.35). CONCLUSION: This study demonstrated that periodontitis and dementia are associated, and that periodontitis is a risk factor for dementia in adults. The risk of dementia was dependent on the periodontal health status of adults, and our findings suggest that regular scaling can reduce the incidence of dementia in adults. Therefore, regular and routine scaling treatment is suggested for adults.

Maternal autoimmune disease and risk of offspring autism spectrum disorder - a nationwide population-based cohort study.

Introduction: Autism spectrum disorder (ASD) is a group of neurodevelopmental disorders which cause long term social and behavior impairment, and its prevalence is on the rise. Studies about the association between maternal autoimmune diseases and offspring ASD have controversial results. The aim of this study was to investigate whether maternal autoimmune diseases increase the risk of ASD in offspring from a population-based perspective. Methods: The data sources were Taiwan's National Health Insurance Research Database (NHIRD) and Taiwan's Maternal and Child Health Database (MCHD), which were integrated and used to identify newborns whose mothers were diagnosed with autoimmune disease. Newborns were matched by maternal age, neonatal gender, and date of birth with controls whose mothers were without autoimmune disease using a ratio of 1:4 between 2004 and 2019. Data on diagnoses of autoimmune disease and autism spectrum disorders were retrieved from NHIRD. Patients who had at least 3 outpatient visits or at least 1 admission with a diagnosis of autoimmune disease and autism spectrum disorders were defined as incidence cases. The risks of ASD in offspring were compared between mothers with or without autoimmune disorders. Results: We identified 20,865 newborns whose mothers had been diagnosed with autoimmune disease before pregnancy and matched them at a ratio of 1:4 with a total of 83,460 newborn whose mothers were without autoimmune disease, by maternal age, neonatal gender, and date of birth. They were randomly selected as the control group. The cumulative incidence rates of autism spectrum disorders (ASD) were significantly higher among the offspring of mothers with autoimmune diseases. After adjusting for cofactors, the risk of ASD remained significantly higher in children whose mother had autoimmune diseases. Regarding to specific maternal autoimmune disease, Sjögren's syndrome and rheumatoid arthritis were both associated with elevated risks of ASD in offspring. Conclusion: Mother with autoimmune disease might be associated with increasing the risk of autism spectrum disorder in offspring.

Handgrip strength is associated with mortality in community-dwelling older adults: the Yilan cohort study, Taiwan.

INTRODUCTION: Hand grip strength (HGS) is one of the methods to help early identification of physical frailty and sarcopenia, the major concerns in the aging societies. It is also crucial to evaluate its impact on mortality. However, the available evidence regarding such impact among specific age cohorts (65 to 74 years and above) is limited. This study tried to investigate the relationship between HGS and mortality among specific cohorts of the community-dwelling older individuals in Yilan, Taiwan. METHODS: A seven-year longitudinal follow-up study was conducted involving 2,468 community-dwelling older individuals in Yilan. The participants were divided into two groups based on their quartiles of hand grip strength: with poor HGS and with good HGS. The association between HGS and mortality was examined using Cox proportional hazards models. RESULTS: The analysis revealed that age, HGS, gender, medical history of cardiovascular diseases, body mass index, and wrist-hip ratio had significant impacts on seven-year survival. Specifically, individuals with poor HGS exhibited increased mortality, with an adjusted hazard ratio (HR) of 1.87 (95% CI: 1.52-2.30). Furthermore, the adverse effect of poor HGS on mortality was more pronounced in males aged 65-74 years (adjusted HR 4.12, 95% CI: 2.16-7.84), females aged 75 years or older (2.09, 1.43-3.04) and males aged 75 years or older (1.49, 1.07-2.07). CONCLUSION: Poor hand grip strength is an independent risk factor for mid-term mortality among community-dwelling older individuals in Yilan. The assessment of HGS can serve as a valuable tool in identifying older individuals at higher risk of death.