Enhancing early detection of cognitive decline in the elderly: a comparative study utilizing large language models in clinical notes.

BACKGROUND: Large language models (LLMs) have shown promising performance in various healthcare domains, but their effectiveness in identifying specific clinical conditions in real medical records is less explored. This study evaluates LLMs for detecting signs of cognitive decline in real electronic health record (EHR) clinical notes, comparing their error profiles with traditional models. The insights gained will inform strategies for performance enhancement. METHODS: This study, conducted at Mass General Brigham in Boston, MA, analysed clinical notes from the four years prior to a 2019 diagnosis of mild cognitive impairment in patients aged 50 and older. We developed prompts for two LLMs, Llama 2 and GPT-4, on Health Insurance Portability and Accountability Act (HIPAA)-compliant cloud-computing platforms using multiple approaches (e.g., hard prompting, retrieval augmented generation, and error analysis-based instructions) to select the optimal LLM-based method. Baseline models included a hierarchical attention-based neural network and XGBoost. Subsequently, we constructed an ensemble of the three models using a majority vote approach. Confusion-matrix-based scores were used for model evaluation. FINDINGS: We used a randomly annotated sample of 4949 note sections from 1969 patients (women: 1046 [53.1%]; age: mean, 76.0 [SD, 13.3] years), filtered with keywords related to cognitive functions, for model development. For testing, a random annotated sample of 1996 note sections from 1161 patients (women: 619 [53.3%]; age: mean, 76.5 [SD, 10.2] years) without keyword filtering was utilised. GPT-4 demonstrated superior accuracy and efficiency compared to Llama 2, but did not outperform traditional models. The ensemble model outperformed the individual models in terms of all evaluation metrics with statistical significance (p < 0.01), achieving a precision of 90.2% [95% CI: 81.9%-96.8%], a recall of 94.2% [95% CI: 87.9%-98.7%], and an F1-score of 92.1% [95% CI: 86.8%-96.4%]. Notably, the ensemble model showed a significant improvement in precision, increasing from a range of 70%-79% to above 90%, compared to the best-performing single model. Error analysis revealed that 63 samples were incorrectly predicted by at least one model; however, only 2 cases (3.2%) were mutual errors across all models, indicating diverse error profiles among them. INTERPRETATION: LLMs and traditional machine learning models trained using local EHR data exhibited diverse error profiles. The ensemble of these models was found to be complementary, enhancing diagnostic performance. Future research should investigate integrating LLMs with smaller, localised models and incorporating medical data and domain knowledge to enhance performance on specific tasks. FUNDING: This research was supported by the National Institute on Aging grants (R44AG081006, R01AG080429) and National Library of Medicine grant (R01LM014239).

Generative Large Language Models in Electronic Health Records for Patient Care Since 2023: A Systematic Review.

Background: Generative Large language models (LLMs) represent a significant advancement in natural language processing, achieving state-of-the-art performance across various tasks. However, their application in clinical settings using real electronic health records (EHRs) is still rare and presents numerous challenges. Objective: This study aims to systematically review the use of generative LLMs, and the effectiveness of relevant techniques in patient care-related topics involving EHRs, summarize the challenges faced, and suggest future directions. Methods: A Boolean search for peer-reviewed articles was conducted on May 19th, 2024 using PubMed and Web of Science to include research articles published since 2023, which was one month after the release of ChatGPT. The search results were deduplicated. Multiple reviewers, including biomedical informaticians, computer scientists, and a physician, screened the publications for eligibility and conducted data extraction. Only studies utilizing generative LLMs to analyze real EHR data were included. We summarized the use of prompt engineering, fine-tuning, multimodal EHR data, and evaluation matrices. Additionally, we identified current challenges in applying LLMs in clinical settings as reported by the included studies and proposed future directions. Results: The initial search identified 6,328 unique studies, with 76 studies included after eligibility screening. Of these, 67 studies (88.2%) employed zero-shot prompting, five of them reported 100% accuracy on five specific clinical tasks. Nine studies used advanced prompting strategies; four tested these strategies experimentally, finding that prompt engineering improved performance, with one study noting a non-linear relationship between the number of examples in a prompt and performance improvement. Eight studies explored fine-tuning generative LLMs, all reported performance improvements on specific tasks, but three of them noted potential performance degradation after fine-tuning on certain tasks. Only two studies utilized multimodal data, which improved LLM-based decision-making and enabled accurate rare disease diagnosis and prognosis. The studies employed 55 different evaluation metrics for 22 purposes, such as correctness, completeness, and conciseness. Two studies investigated LLM bias, with one detecting no bias and the other finding that male patients received more appropriate clinical decision-making suggestions. Six studies identified hallucinations, such as fabricating patient names in structured thyroid ultrasound reports. Additional challenges included but were not limited to the impersonal tone of LLM consultations, which made patients uncomfortable, and the difficulty patients had in understanding LLM responses. Conclusion: Our review indicates that few studies have employed advanced computational techniques to enhance LLM performance. The diverse evaluation metrics used highlight the need for standardization. LLMs currently cannot replace physicians due to challenges such as bias, hallucinations, and impersonal responses.

Ultrasound attenuation imaging as a strategy for evaluation of early and late ambulatory functions in Duchenne muscular dystrophy.

BACKGROUND: Duchenne muscular dystrophy (DMD) is a genetic neuromuscular disorder that leads to mobility loss and life-threatening cardiac or respiratory complications. Quantitative ultrasound (QUS) envelope statistics imaging, which characterizes fat infiltration and fibrosis in muscles, has been extensively used for DMD evaluations. PURPOSE: Notably, changes in muscle microstructures also result in acoustic attenuation, potentially serving as another crucial imaging biomarker for DMD. Expanding upon the reference frequency method (RFM), this study contributes to the field by introducing the robust RFM (RRFM) as a novel approach for ultrasound attenuation imaging in DMD. METHODS: The RRFM algorithm was developed using an iterative reweighted least squares technique. We conducted standard phantom measurements with a clinical ultrasound system equipped with a linear array transducer to assess the improvement in attenuation estimation bias by RRFM. Additionally, 161 DMD patients, included in both a validation dataset (n = 130) and a testing dataset (n = 31), underwent ultrasound scanning of the gastrocnemius for RRFM-based attenuation imaging. The diagnostic performances for ambulatory functions and discrimination between early and late ambulatory stages were evaluated and compared with those of QUS envelope statistics imaging (involving Nakagami distribution, homodyned K distribution, and entropy values) using the area under the receiver operating characteristic curve (AUROC). RESULTS: The results indicated that the RRFM method more closely matched the actual attenuation properties of the phantom, reducing measurement bias by 50% compared to conventional RFM. The AUROCs for RRFM-based attenuation imaging, used to discriminate between early and late ambulatory stages, were 0.88 and 0.92 for the validation and testing datasets, respectively. These performances significantly surpassed those of QUS envelope statistics imaging (p < 0.05). CONCLUSIONS: Ultrasound attenuation imaging employing RRFM may serve as a sensitive tool for evaluating the progression of ambulatory function deterioration, offering substantial potential for the health management and follow-up care of DMD patients.

Risk of new onset hyperuricemia and chronic kidney disease after living kidney donation through propensity score matching analysis.

Living kidney donors have been regarded as those people having earned the healthiest status level after having undergone scrutiny. Although one's post-donation GFR is expected to fall to 50% of their pre-donation value, it is well documented that there is a compensatory increase in GFR which subsequently reaches approximately 60-70% of the donor's pre-donation value. Data regarding gout/hyperuricemia in living kidney donors has remained scarce until now. This study involved kidney donors enrolled within the years 2000 to 2017, where those who were selected to be matched to those in group of case cohort by age, year of index date, gender and co-morbidity were considered as the control cohort. During the 17-year study period 2,716 participants were enrolled. Results revealed that kidney donors experienced a risk of new onset gout/ hyperuricemia (adjusted HR = 1.73; 95%CI = 1.27, 2.36), and new onset CKD (adjusted HR = 6.7; 95% CI = 4.4, 10.21) were found to be higher in kidney donors. Our findings suggest that people after kidney donation are significantly associated with a higher risk of new onset gout/hyperuricemia. Clinical professionals therefore need to be cautious of new onset gouy/hyperuricemia after donation surgery.

Enhancing Early Detection of Cognitive Decline in the Elderly: A Comparative Study Utilizing Large Language Models in Clinical Notes.

Background: Large language models (LLMs) have shown promising performance in various healthcare domains, but their effectiveness in identifying specific clinical conditions in real medical records is less explored. This study evaluates LLMs for detecting signs of cognitive decline in real electronic health record (EHR) clinical notes, comparing their error profiles with traditional models. The insights gained will inform strategies for performance enhancement. Methods: This study, conducted at Mass General Brigham in Boston, MA, analyzed clinical notes from the four years prior to a 2019 diagnosis of mild cognitive impairment in patients aged 50 and older. We used a randomly annotated sample of 4,949 note sections, filtered with keywords related to cognitive functions, for model development. For testing, a random annotated sample of 1,996 note sections without keyword filtering was utilized. We developed prompts for two LLMs, Llama 2 and GPT-4, on HIPAA-compliant cloud-computing platforms using multiple approaches (e.g., both hard and soft prompting and error analysis-based instructions) to select the optimal LLM-based method. Baseline models included a hierarchical attention-based neural network and XGBoost. Subsequently, we constructed an ensemble of the three models using a majority vote approach. Results: GPT-4 demonstrated superior accuracy and efficiency compared to Llama 2, but did not outperform traditional models. The ensemble model outperformed the individual models, achieving a precision of 90.3%, a recall of 94.2%, and an F1-score of 92.2%. Notably, the ensemble model showed a significant improvement in precision, increasing from a range of 70%-79% to above 90%, compared to the best-performing single model. Error analysis revealed that 63 samples were incorrectly predicted by at least one model; however, only 2 cases (3.2%) were mutual errors across all models, indicating diverse error profiles among them. Conclusions: LLMs and traditional machine learning models trained using local EHR data exhibited diverse error profiles. The ensemble of these models was found to be complementary, enhancing diagnostic performance. Future research should investigate integrating LLMs with smaller, localized models and incorporating medical data and domain knowledge to enhance performance on specific tasks.

Deep learning-based glomerulus detection and classification with generative morphology augmentation in renal pathology images.

Glomerulus morphology on renal pathology images provides valuable diagnosis and outcome prediction information. To provide better care, an efficient, standardized, and scalable method is urgently needed to optimize the time-consuming and labor-intensive interpretation process by renal pathologists. This paper proposes a deep convolutional neural network (CNN)-based approach to automatically detect and classify glomeruli with different stains in renal pathology images. In the glomerulus detection stage, this paper proposes a flattened Xception with a feature pyramid network (FX-FPN). The FX-FPN is employed as a backbone in the framework of faster region-based CNN to improve glomerulus detection performance. In the classification stage, this paper considers classifications of five glomerulus morphologies using a flattened Xception classifier. To endow the classifier with higher discriminability, this paper proposes a generative data augmentation approach for patch-based glomerulus morphology augmentation. New glomerulus patches of different morphologies are generated for data augmentation through the cycle-consistent generative adversarial network (CycleGAN). The single detection model shows the F1 score up to 0.9524 in H&E and PAS stains. The classification result shows that the average sensitivity and specificity are 0.7077 and 0.9316, respectively, by using the flattened Xception with the original training data. The sensitivity and specificity increase to 0.7623 and 0.9443, respectively, by using the generative data augmentation. Comparisons with different deep CNN models show the effectiveness and superiority of the proposed approach.

Association of urinary liver-type fatty acid-binding protein with outcomes after adrenalectomy for unilateral primary aldosteronism.

BACKGROUND AND AIMS: This study explored the association between urinary liver-type fatty acid-binding protein to creatinine (uL-FABP-cre) ratio and postoperative clinical failure in unilateral primary aldosteronism (PA) patients undergoing adrenalectomy. MATERIALS AND METHODS: Data from the Taiwan Primary Aldosteronism Investigation Group database were analyzed, including patients with unilateral PA who had adrenalectomy between December 2015 and October 2018. Statistical methods included generalized additive modeling, logistic regression analysis, net reclassification improvement (NRI), and the C statistic. RESULTS: In the study cohort of 131 patients (mean age 52.3 ± 10.8 years; 43.5% male), 117 achieved clinical success, while 14 experienced clinical failure. A uL-FABP-cre ratio ≥5 predicted clinical failure (odds ratio: 6.22, p = 0.005). Subgroup analysis revealed its efficacy in predicting clinical failure in patients with BMI ≥ 24 kg/m2, normokalemia, or <5 years of hypertension. Furthermore, incorporating uL-FABP-cre ratio into the Primary Aldosteronism Surgical Outcome (PASO) score significantly improved predictive ability. The addition increased the C statistic from 0.671 to 0.762 (p < 0.01) and improved category-free NRI by 0.675 (p = 0.014). CONCLUSION: A uL-FABP-cre ratio ≥5 accurately predicted clinical failure post-adrenalectomy in unilateral PA, enhancing PASO score's identification of high-risk patients for postoperative clinical failure.

Diagnosis and Management of Fabry Disease in High-Risk Renal Disease Patients in Taiwan: A Single Center Study.

BACKGROUND: Fabry disease (FD) is an X-linked inborn error of lysosomal storage disorder, a deficiency in lysosomal hydrolase α-galactosidase A activity due to pathogenic variants in the GLA gene. Accumulation of globotriaosylceramide in multiple organs contributes to end-stage kidney disease, heart failure, and cerebrovascular accidents. METHODS: We began the FD screening program by involving male patients older than 20 years of age who were on chronic dialysis, had a post-kidney transplantation, and were part of the Pre-End Stage Renal Disease Program in our hospital. α-galactosidase A activity was detected through an initial dried blood spots screen assay, followed by levels of lyso-globotriaosylceramide and sequencing of the GLA gene when screening patients with suspected FD to confirm their diagnosis. RESULTS: A total of 1812 patients had been FD screened, with the prevalence of FD being approximately 0.16 % (3/1812) up until June 2022. Interestingly, we confirmed a family cluster (2 sons and their mother) of having the c.936+919G>A mutation (designated GLA IVS4) with hypertrophic cardiomyopathy in Taiwan and another with the mutation c.644A>G (p.Asn215Ser), a more common later-onset variant reported in people of European or North American descent. Two patients were confirmed with cardiomyopathy through a cardiac biopsy, with their cardiac function later reversed after enzyme replacement therapy. CONCLUSIONS: The FD screening test detects chronic kidney disease due to an unknown etiology and prevents other organ complications. Early detection of FD is crucial for reversing target organ damage with enzyme replacement therapy.

Outcomes and Effects of Vaccination on Sars-Cov-2 Omicron Infection in Kidney Transplant Recipients.

BACKGROUND: Kidney transplant recipients (KTRs) are at high risk of COVID-19. Vaccination is significantly effective at preventing infection and reducing infection severity. Omicron infections are less severe than infections by previous strains, but breakthrough disease is more common. Thus, we conducted this study to observe the vaccine efficacy in our KTRs. METHODS: During the surge in the Omicron variant, beginning in May 2022, we retrieved data from 365 KTRs who had received at least one dose of various COVID vaccines until June 30, 2022. Outcomes of the KTRs (n = 168) after at least the 2nd vaccination were assessed until September 30, 2022, before the border was opened for tourism. RESULTS: The antibody response in KTRs after the 1st and 2nd doses of SARS-CoV-2 vaccines demonstrated a significant increase from the 1st dose (median: 0.4; IQR: 0.4-8.4 U/mL, P < .001) to the 2nd dose (median: 57.5; IQR: 0.4-799.2 U/mL), and the response rate rose from 32% to 65% (P < .001). SARS-CoV-2 infection was identified in 14/365 (3.8%) patients after at least the 1st dose and 7/187 (3.7%) patients at least 7 days beyond the 2nd dose. Most KTRs had a mild course, but 3 (17%) were hospitalized due to pneumonia. CONCLUSIONS: Our data demonstrate a lower response rate and anti-S titers after 2nd dose vaccination in KTRs than in the general population, but a lower incidence of SARS-CoV-2 infection after vaccination was observed during the Omicron outbreak. Owing to the breakthrough infections found in ordinarily vaccinated KTRs, we must emphasize the importance of vaccination and boosters to prevent severe illness, hospitalizations, and death among those developing infections.

Adverse Effects of Denosumab in Kidney Transplant Recipients: A 20-Year Retrospective Single-Center Observation Study in Central Taiwan.

BACKGROUND: Bone loss can be noted in kidney transplantation recipients (KTRs) and can be related to fracture events. Denosumab, a potent monoclonal antibody to RANK ligand, increases lumbar bone mineral density. However, safety data for denosumab remain limited regarding transplant recipients. Hypocalcemia and increased genital tract infections have been mentioned as adverse effects in KTRs after being prescribed denosumab. METHODS: We retrospectively analyzed the electronic medical records of KTRs during the recent 20 years who had been prescribed antiresorptive therapy and were >18 years old. Medical records and their clinical data were reviewed and analyzed. We compared the frequency of adverse effects between denosumab with other antiresorptive therapies. RESULTS: A total of 70 KTRs were enrolled, with 46 patients being given denosumab and the first injection being noted on October 31, 2014. No significant differences were seen in mortality rate, opportunistic infection, pneumonia, or genitourinary tract infection. One diagnosis of osteonecrosis of the jaw was noted in the denosumab group (2.2%). A higher incidence of hypocalcemia (<8.4 mg/dL) was noted in the denosumab group (34.8%), and a higher but nonsignificant difference in the incidence of severe hypocalcemia was also noted in the group. CONCLUSIONS: Denosumab may be considered as safe as other antiresorptive therapies for KTRs. However, more hypocalcemia events have been noted, so medical personnel may need to be cautious when prescribing its use.

Outcome of Brain Lymphoma in a High Epstein-Barr Virus-Prevalence Country After Kidney Transplantation.

BACKGROUND: The incidence of post-transplant lymphoproliferative disorder (PTLD) in adult kidney transplant (KTx) recipients is less common in Taiwan. In our institute, we observed that brain lymphoma was the most notorious type. METHODS: The study describes the clinical, histologic, and radiological features of primary central nervous lymphoma (PCNSL) and the outcomes and associations with Epstein-Barr virus (EBV) infection in our center. RESULTS: Among 1470 KTx recipients, 5 patients had tissue-proven brain lymphoma (0.34%). The brain pathology disclosed diffuse large B-cell lymphoma in all patients. EBV was detected through in situ hybridization for Epstein-Barr encoding region (EBER) to disclose the EBV inclusion in the nuclei of lymphoma cells. The first treatment step was the reduction of immunosuppressants; 4 patients received whole-brain radiotherapy after complete resection of PCNSL, and 1 received concurrent chemoradiotherapy. Only one patient had poor performance status at the time of diagnosis and had a poor response to treatment with steroids. Four patients survived (mean 36.5 months, range 8.6 to 57.6 months), but one died after rapid neurologic deterioration. CONCLUSION: Epstein-Barr virus inclusion was found in PCNSL in our patients; however, the role of EBV in PCNSL remains to be clarified. Post-transplant lymphoproliferative disorder is a rare malignancy after KTx with a predilection of brain involvement in Taiwan. We report a successful care experience in a patient with primary CNS lymphoma with better survival.

Risk of Severe Alphaherpesvirus Infection after Solid Organ Transplantation: A Nationwide Population-Based Cohort Study.

Patients after solid organ transplantation (SOT) are more susceptible to various viral infections, including alphaherpesviruses. Therefore, the aim of our study was to investigate the risk of alphaherpesvirus infections, including herpes simplex and herpes zoster, after solid organ transplantation. Inpatient records from the Taiwan National Health Insurance Research Database (NHIRD) defined solid organ recipients, including heart, liver, lung, and kidney, hospitalized for alphaherpesvirus infections as a severe case group of transplants and matched them with a nontransplant cohort. We enrolled 18,064 individuals, of whom 9032 were in each group. A higher risk of severe alphaherpesvirus infection was noted in solid organ recipients (aHR = 9.19; p < 0.001) than in the general population. In addition, solid organ transplant recipients had the highest risk of alphaherpesvirus infection within 1 year after transplantation (aHR = 25.18). The comparison found a higher risk of herpes zoster and herpes simplex infections in recipients of kidney (aHR = 9.13; aHR = 12.13), heart (aHR = 14.34; aHR = 18.54), and liver (aHR = 5.90; aHR = 8.28) transplants. Patients who underwent solid organ transplantation had a significantly higher risk of alphaherpesvirus infection than the general population.

Risk of Major Cardiovascular Disease after Exposure to Contrast Media: A Nationwide Population-Based Cohort Study on Dialysis Patients.

Contrast associated kidney injury is caused by side effects of iodinated contrast media (ICM), including inflammation. Chronic inflammation among dialysis patient contributes to atherosclerosis, which leads to simultaneous conditions of the kidney, brain, and vasculature. Data to investigate the pathologic effects of ICM on cardiovascular complications in dialysis patients are lacking. Dialysis patients who had been exposed to ICM from computed tomography (ICM-CT) were allocated as the ICM-CT cohort (N = 3751), whereas dialysis patients without ICM exposure were randomly allocated as the non-ICM cohort (N = 17,196). Furthermore, 540 pairs were selected for analyses through propensity score-matching in terms of age, sex, comorbidities, dialysis vintage, and index date. During a median follow-up of 10.3 years, ICM-CT cohort had significantly higher risks in the following, compared with non-ICM cohort: all-cause mortality (adjusted hazard ratio [aHR], 1.36; 95% confidence interval [CI], 1.26-1.47), cardiovascular events (aHR,1.67; 95% CI, 1.39-2.01), acute coronary syndrome (adjusted HR: 2.92; 95% CI, 1.72-4.94), sudden cardiac arrest (aHR, 1.69; 95% CI, 0.90-3.18), heart failure (aHR, 1.71; 95% CI,1.28-2.27), and stroke (aHR, 1.84; 95% CI,1.45-2.35). The proinflammatory ICM is significantly associated with an increased risk of major cardiovascular events in patients on dialysis.

Patterns of biopsy-proven renal diseases in geriatric patients: A single medical center experience.

The elderly population is expanding rapidly, and that has become a major healthcare burden in terms of chronic kidney disease. The distribution patterns of kidney diseases in these elderly patients remain largely unclear. Here, we compared biopsy-based renal disease patterns between elderly and nonelderly patients. We performed a single-center, retrospective study (1992-2008) on biopsy-proven renal diseases to compare results between geriatric patients (age ≥ 65 years; n = 254) and nongeriatric patients (18 ≤ age < 65 years; n = 2592). Renal pathology was interpreted by pathologists based on light microscopy, immunofluorescence, and electron microscopy. The ages of the geriatric and nongeriatric groups were 71.8 ±â€…4.5 (65.1-87.3) and 39.7 ±â€…17.6 (18-64.9) years, respectively, and 74% and 41% of them, respectively, were men. In the geriatric group, the most frequent diagnosis was membranous nephropathy (46.1%), followed by minimal change disease/focal segmental glomerulosclerosis (16.9%), diabetic nephropathy (8.3%), hypertensive nephrosclerosis (7.5%), and IgA nephropathy (5.9%). The geriatric group had more membranous nephropathy and less lupus nephritis and IgA nephropathy than the nongeriatric group. Furthermore, the 5-year survival rate of the geriatric group was significantly low. Our results demonstrated the different distributions of renal biopsy patterns in geriatric patients diagnosed with acute or chronic progressive kidney injury and proteinuria through renal biopsy.

Gender Differences in microRNA Expressions as Related to Long-Term Graft Function in Kidney Transplant Patients.

In recent studies, much has been discussed about biomarkers used in the evaluation of the transplanted graft function. However, there remains a lack of research regarding the long-term effects of microRNAs (miRNAs) on the different genders for kidney transplant (KTx) patients. In this study, we aim to assess the functions of miRNAs on long term outcomes of KTx patients by extracting differently expressed miRNAs between patients of normal graft function and graft dysfunction, while further analyzing their impact on the different genders. We analyzed the data of 40 patients who had received KTx for a period of more than ten years and included data regarding renal function, immuno-related markers and plasma miRNAs. Data were classified by gender for further studies. Twelve out of 17 females and 8 out of 23 males had undergone graft dysfunction. Renal function analysis showed significantly worse outcomes in the female patients. There were five differently expressed miRNAs between the female control group and female dysfunction group: miR-128-3p, miR-21-5p, miR-150-5p, miR-92a-3p and miR-15a-5p, and five between the male control group and male dysfunction group: miR-23a-3p, miR-126-3p, miR-142-3p, miR-223-3p and miR-26a-5p. Gender differences exist in incidences of kidney graft dysfunction, with male patients displaying better preservation in graft functions. Overall, these differently expressed miRNAs either enhance or suppress host immune responses. They can be predictive markers for graft survival and can also be important factors that lead to worse long term kidney graft function in females when compared to males.

Risk of Dementia after Exposure to Contrast Media: A Nationwide, Population-Based Cohort Study.

Contrast-medium-associated kidney injury is caused by the infusion of contrast media. Small vessel disease is significantly associated with various diseases, including simultaneous conditions of the kidney and brain, which are highly vulnerable to similar vascular damage and microvascular pathologies. Data to investigate the adverse effect of contrast media on the brain remain extremely lacking. In this study, 11,332,616 NHI enrollees were selected and divided into two groups, exposed and not exposed to a contrast medium during the observation period, from which 1,461,684 pairs were selected for analyses through matching in terms of age, sex, comorbidities, and frequency of outpatient visits during the previous year. In total, 1,461,684 patients exposed to a contrast medium and 1,461,684 controls not exposed to one were enrolled. In multivariable Cox proportional hazard models, patients exposed to a contrast medium had an overall 2.09-fold higher risk of dementia. In multivariable-stratified analyses, the risk of Alzheimer's disease was remarkably high in younger patients without any underlying comorbidity. This study is the first to discover that exposure to contrast media is significantly associated with the risk of dementia. A four-fold increased risk of vascular dementia was observed after exposure to a contrast medium. Further studies on the influence of exposure to contrast media on the brain are warranted.

Early Comprehensive Kidney Care in Dialysis-Requiring Acute Kidney Injury Survivors: A Populational Study.

Background: For patients with Acute Kidney Injury (AKI), a strong and graded relationship exists between AKI severity and mortality. One of the most severe entities of AKI is Dialysis-Requiring Acute Kidney Injury (D-AKI), which is associated with high rates of mortality and end-stage renal disease (ESRD). For this high-risk population group, there is a lack of evidence regarding optimal post-AKI care. We propose that post-AKI care through the combined efforts of the nephrologist and the multidisciplinary care team may improve outcomes. Our aim here is to study for survivors of dialysis-requiring acute kidney injury, the effects of implementing early comprehensive kidney care. Methods: This is a retrospective longitudinal cohort study of Taiwanese through analyzing the National Health Insurance Research Database (NHIRD). We included patients with acute dialysis during hospitalization from January 1, 2015 to December 31, 2018. Propensity match was done at 1:1, including estimated glomerular filtration rate (eGFR) based on CKD-EPI which was performed due to large initial disparities between these two cohorts. Results: After the propensity match, each cohort had 4,988 patients. The mean eGFR based on CKD-EPI was 27.5 ml/min/1.73 m2, and the mean follow-up period was 1.4 years.The hazard ratio for chronic dialysis or ESRD was 0.55 (95% CI, 0.49-0.62; p < 0.001). The hazard ratio for all-cause mortality was 0.79 (95% CI, 0.57-0.88; p < 0.001). Both outcomes favored early comprehensive kidney care. Conclusions: For survivors of dialysis-requiring acute kidney injury, early comprehensive kidney care significantly lowered risks of chronic dialysis and all-cause mortality.

Impact of peritoneal dialysis-related peritonitis on PD discontinuation and mortality: A population-based national cohort study.

BACKGROUND: The impact of peritoneal dialysis-associated peritonitis (PD peritonitis) on long-term outcomes is uncertain. This nationwide retrospective study was conducted in Taiwan to understand the incidence, risk factors and long-term outcomes of PD peritonitis. METHODS: A total of 11,202 incident adult peritoneal dialysis (PD) patients from 2000 to 2010 were collected from a Longitudinal Health Insurance Database and followed up until the end of 2011. Definition of peritonitis, the primary outcome, simultaneously met the diagnosis of peritonitis (International Classification of Diseases, Ninth Revision, Clinical Modification 567) and antibiotic use. Secondary outcomes included the impact of peritonitis on PD discontinuation and survival. Cox proportional hazards models with and without time-dependent variables were used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs). RESULTS: There were 7634 peritonitis episodes in 4245 patients during the follow-up period. The overall incidence of peritonitis was 0.18 episodes per patient-year. Peritonitis-associated risk factors included older age, female gender, chronic heart failure, cerebrovascular disease, liver cirrhosis and lower monthly income. In an adjusted Cox hazard proportional regression with the time-dependent model, peritonitis patients had a higher risk of PD discontinuation (HR 2.71, 95% CI 2.52-2.92) and mortality (HR 1.68, 95% CI 1.57-1.81) compared to patients without peritonitis. The adjusted HRs for mortality increased with each prior episode: one episode, two episodes and more than two episodes (all p < 0.05). The adjusted HRs for PD discontinuation also increased with the frequency of peritonitis. These negative effects were greatest during the first year and persisted significantly after 5 years. In a sensitivity analysis in which peritonitis within 30 days of death or PD discontinuation was excluded, peritonitis patients still had significantly increased risk of PD discontinuation and mortality compared to patients without peritonitis. CONCLUSIONS: Although peritonitis incidence was low, our findings reveal that peritonitis carried acute and long-term sequelae of higher PD discontinuation and lower patient survival.

Arteriovenous Fistula Flow Dysfunction Surveillance: Early Detection Using Pulse Radar Sensor and Machine Learning Classification.

Vascular Access (VA) is often referred to as the "Achilles heel" for a Hemodialysis (HD)-dependent patient. Both the patent and sufficient VA provide adequacy for performing dialysis and reducing dialysis-related complications, while on the contrary, insufficient VA is the main reason for recurrent hospitalizations, high morbidity, and high mortality in HD patients. A non-invasive Vascular Wall Motion (VWM) monitoring system, made up of a pulse radar sensor and Support Vector Machine (SVM) classification algorithm, has been developed to detect access flow dysfunction in Arteriovenous Fistula (AVF). The harmonic ratios derived from the Fast Fourier Transform (FFT) spectrum-based signal processing technique were employed as the input features for the SVM classifier. The result of a pilot clinical trial showed that a more accurate prediction of AVF flow dysfunction could be achieved by the VWM monitor as compared with the Ultrasound Dilution (UD) flow monitor. Receiver Operating Characteristic (ROC) curve analysis showed that the SVM classification algorithm achieved a detection specificity of 100% at detection thresholds in the range from 500 to 750 mL/min and a maximum sensitivity of 95.2% at a detection threshold of 750 mL/min.

Evaluating the renal safety of tenofovir disoproxil fumarate in hepatitis B patients without chronic kidney disease.

The nephrotoxicity of tenofovir disoproxil fumarate (TDF) in chronic hepatitis B (CHB) patients without chronic kidney disease (CKD) remains controversial. We aimed to evaluate nephrotoxicity of TDF in this population. In this hospital-based cohort study, CHB patients who received either TDF or entecavir (ETV) therapy, and did not have underlying CKD, were retrospectively recruited from January, 2008 to January, 2019. After excluding those with confounding conditions, 257 TDF-treated patients were matched through propensity scores with 514 ETV-treated patients. Cumulative incidences of, and hazard ratios (HRs) for the CKD guideline-defined renal dysfunction, were analysed. The mean decline in glomerular filtration rate was similar over 60 months (TDF vs. ETV: 10.1 ml/min/1.73 m2 , 95% confidence interval [CI]: 7.4-12.7 vs. 8.0 ml/min/1.73 m2 , 95% CI: 6.4-9.6; p = .34). The 5-year cumulative incidence of renal dysfunction was not significantly different (TDF vs. ETV: 10.4%, 95% CI: 5.6-18.0 vs. 5.8%, 95% CI: 3.6-9.0; p = .18). However, in multivariable stratified analysis, TDF was associated with an increased risk of renal dysfunction in the elderly (age ≥60 years), when compared to ETV (HR 2.86, 95% CI: 1.02-8.01; p < .05). For confirming the effect of TDF amongst the elderly, 61 TDF-treated patients were further matched with 183 ETV-treated patients, with 5-year cumulative incidence of renal dysfunction being significantly higher in TDF users (TDF vs. ETV: 34.4%, 95% CI: 17.7-59.8 vs. 15.5%, 95% CI: 9.4-25.1; p < .05). TDF use was independently related to renal dysfunction (HR 2.71, 95% CI: 1.19-6.14; p < .05). Although TDF is generally safe for CHB patients without CKD, it is best to be avoided in the elderly.