Clustering using unsupervised machine learning to stratify the risk of immune-related liver injury.

BACKGROUND AND AIM: Immune-related liver injury (liver-irAE) is a clinical problem with a potentially poor prognosis. METHODS: We retrospectively collected clinical data from patients treated with immune checkpoint inhibitors between September 2014 and December 2021 at the Nagoya University Hospital. Using an unsupervised machine learning method, the Gaussian mixture model, to divide the cohort into clusters based on inflammatory markers, we investigated the cumulative incidence of liver-irAEs in these clusters. RESULTS: This study included a total of 702 patients. Among them, 492 (70.1%) patients were male, and the mean age was 66.6 years. During the mean follow-up period of 423 days, severe liver-irAEs (Common Terminology Criteria for Adverse Events grade ≥ 3) occurred in 43 patients. Patients were divided into five clusters (a, b, c, d, and e). The cumulative incidence of liver-irAE was higher in cluster c than in cluster a (hazard ratio [HR]: 13.59, 95% confidence interval [CI]: 1.70-108.76, P = 0.014), and overall survival was worse in clusters c and d than in cluster a (HR: 2.83, 95% CI: 1.77-4.50, P < 0.001; HR: 2.87, 95% CI: 1.47-5.60, P = 0.002, respectively). Clusters c and d were characterized by high temperature, C-reactive protein, platelets, and low albumin. However, there were differences in the prevalence of neutrophil count, neutrophil-to-lymphocyte ratio, and liver metastases between both clusters. CONCLUSIONS: The combined assessment of multiple markers and body temperature may help stratify high-risk groups for developing liver-irAE.

Identification of clinical phenotypes associated with poor prognosis in patients with nonalcoholic fatty liver disease via unsupervised machine learning.

BACKGROUND AND AIMS: Both fibrosis status and body weight are important for assessing prognosis in nonalcoholic fatty liver disease (NAFLD). The aim of this study was to identify population clusters for specific clinical outcomes based on fibrosis-4 (FIB-4) index and body mass index (BMI) using an unsupervised machine learning method. METHODS: We conducted a multicenter study of 1335 biopsy-proven NAFLD patients from Japan. Using the Gaussian mixture model to divide the cohort into clusters based on FIB-4 index and BMI, we investigated prognosis for these clusters. RESULTS: The cohort consisted of 223 cases (16.0%) with advanced fibrosis (F3-4) as assessed from liver biopsy. Median values of BMI and FIB-4 index were 27.3 kg/m2 and 1.67. The patients were divided into four clusters by Bayesian information criterion, and all-cause mortality was highest in cluster d, followed by cluster b (P = 0.001). Regarding the characteristics of each cluster, clusters d and b presented a high FIB-4 index (median 5.23 and 2.23), cluster a presented the lowest FIB-4 index (median 0.78), and cluster c was associated with moderate FIB-4 level (median 1.30) and highest BMI (median 34.3 kg/m2 ). Clusters a and c had lower mortality rates than clusters b and d. However, all-cause of death in clusters a and c was unrelated to liver disease. CONCLUSIONS: Our clustering approach found that the FIB-4 index is an important predictor of mortality in NAFLD patients regardless of BMI. Additionally, non-liver-related diseases were identified as the causes of death in NAFLD patients with low FIB-4 index.

Abnormal magnesium levels and their impact on death and acute kidney injury in critically ill children.

BACKGROUND: The prevalence of magnesium imbalance in critically ill children is very high. However, its significance in the development of acute kidney injury (AKI) and mortality remains unknown. METHODS: In this retrospective observational study from 2010 to 2018, the pediatric-specific intensive care database was analyzed. We included critically ill children aged > 3 months and those without chronic kidney disease. Patients were diagnosed with AKI, according to the Kidney Disease Improving Global Outcomes (KDIGO) study. We calculated the initial corrected magnesium levels (cMg) within 24 h and used a spline regression model to evaluate the cut-off values for cMg. We analyzed 28-day mortality and its association with AKI. The interaction between AKI and magnesium imbalance was evaluated. RESULTS: The study included 3,669 children, of whom 105 died within 28 days, while 1,823 were diagnosed with AKI. The cut-off values for cMg were 0.72 and 0.94 mmol/L. Both hypermagnesemia and hypomagnesemia were associated with 28-day mortality (odds ratio [OR] = 2.99, 95% confidence interval [CI] = 1.89-4.71, p < 0.001; OR = 2.80, 95% CI = 1.60-4.89, p < 0.001). Hypermagnesemia was associated with AKI (OR = 1.52, 95% CI = 1.27-1.82, p < 0.001), while neither hypermagnesemia nor hypomagnesemia interacted with the AKI stage on the 28-day mortality. CONCLUSIONS: Abnormal magnesium levels were associated with 28-day mortality in critically ill children. AKI and hypermagnesemia had a strong association. "A higher resolution version of the Graphical abstract is available as Supplementary information".

Influence of COVID-19 on the 10-year carbon footprint of the Nagoya University Hospital and medical research centre.

BACKGROUND: Amidst the climate crisis, a key goal of the medical sector is to reduce its large carbon footprint. Although the Coronavirus disease 2019 (COVID-19) pandemic greatly impacted the medical sector, its influence on carbon footprints remains unknown. Therefore, the aim of this study was to evaluate changes in the carbon footprint of a university hospital with a medical research centre over the past 10 years. METHODS: Data on electricity, gas, and water usage, pharmaceutical and medical supply costs, and waste amounts were recorded for Nagoya University Hospital from April 2010 to March 2021. The relevant emission factors were obtained from the Japanese government and the overall monthly carbon footprint was reported according to the Greenhouse Gas Protocol. The effect of the COVID-19 pandemic on the carbon footprint was then compared for three types of emission sources. Moreover, a regression model was used to plot quadratic functions as approximate functions using monthly carbon emissions and monthly average external temperatures. Finally, the monthly carbon footprint was calculated per hospital admission. RESULTS: The overall carbon footprint of the hospital was 73,546 tCO2e in 2020, revealing an increase of 26.60% over the last 10 years. Carbon emissions from electricity consumption represented 26% of total emissions. The individual carbon footprints of pharmaceuticals, medical supplies, waste, and water usage also increased from 2010 to 2020. The overall monthly carbon footprint was positively correlated with the average monthly temperature (R2 = 0.7566, p < 0.001). Compared with 2019, the overall carbon footprint decreased by 2.19% in 2020. Moreover, the monthly carbon footprint per hospital admission increased significantly between 2018 (0.24 tCO2e/admission) and 2020 (0.26 tCO2e/admission) (p = 0.002). CONCLUSION: The overall carbon footprint of the hospital generally increased over the last decade. During the COVID-19 epidemic in 2020, the carbon footprint decreased slightly, likely because of the reduced number of patients. However, the carbon footprint per admission increased, which was attributed to more complicated patient backgrounds because of the ageing population. Therefore, evaluation of carbon emissions in the medical sector is urgently required in order to act on the climate crisis as soon as possible.

Relationship between mortality and use of sodium bicarbonate at the time of dialysis initiation: a prospective observational study.

BACKGROUND: Patients with chronic kidney disease often experience metabolic acidosis. Whether oral sodium bicarbonate can reduce mortality in patients with metabolic acidosis has been debated for years. Hence, this study was conducted to evaluate the utility of sodium bicarbonate in patients who will undergo dialysis therapy. In this study, we investigated the effect of oral sodium bicarbonate therapy on mortality in patients with end-stage kidney disease (ESKD) initiated on dialysis therapy. METHODS: We conducted an observational study of patients when they started dialysis therapy. There were 17 centres participating in the Aichi Cohort Study of Prognosis in Patients Newly Initiated into Dialysis. Data were available on patients' sex, age, use of sodium bicarbonate, drug history, medical history, vital data, and laboratory data. We investigated whether patients on oral sodium bicarbonate for more than three months before dialysis initiation had a better prognosis than those without sodium bicarbonate therapy. The primary outcome was defined as all-cause mortality. RESULTS: The study included 1524 patients with chronic kidney disease who initiated dialysis between October 2011 and September 2013. Among them, 1030 were men and 492 women, with a mean age of 67.5 ± 13.1 years. Of these, 677 used sodium bicarbonate and 845 did not; 13.6% of the patients in the former group and 21.2% of those in the latter group died by March 2015 (p <  0.001). Even after adjusting for various factors, the use of sodium bicarbonate independently reduced mortality (p <  0.001). CONCLUSIONS: The use of oral sodium bicarbonate at the time of dialysis initiation significantly reduced all-cause mortality in patients undergoing dialysis therapy.

Parity, infertility, age at first birth and risk of atrial fibrillation: data from the HUNT study.

AIMS: Although parity, infertility, and age at first birth are important for later development of cardiovascular disease, research on their association with atrial fibrillation (AF) is limited. METHODS: We linked data from the population-based HUNT study, the Medical Birth Registry of Norway (MBRN) and validated medical records from local hospitals. A total of 24,015 women aged 45 years or older were followed for verified incident AF. Parity and age at first birth were retrieved from the MBRN or from self-reported questionnaires in the HUNT. History of infertility was self-reported on the HUNT questionnaire. Cox-proportional hazard models were used to calculate hazard ratios (HR) for the multivariable-adjusted associations of parity, infertility, and age at first birth with risk of AF. RESULTS: During a median follow-up of 12.8 years, 1,448 (6.0%) participants developed AF. Women with higher parity (four or more births vs. two births) were at 21% higher risk of AF (HR 1.21, 95% confidence interval (CI), 1.05-1.39). History of infertility was also associated with risk of AF (HR 1.20, 95% CI, 1.02-1.42). Among parous women, younger age at first birth (<20 years vs. 20-29 years) was associated with a 20% higher risk of AF (HR 1.20, 95% CI, 1.03-1.40). CONCLUSION: Women with four or more births, or a history of infertility, or younger age at first birth have approximately a 20% higher risk of AF among women over 45 years old.

A higher number of births and younger age at first birth are associated with a higher risk of cardiovascular disease (CVD). However, there is limited evidence on the associations between parity, age at first birth and atrial fibrillation (AF). Moreover, the association between infertility and AF remains largely unexplored. We have investigated the association between parity, infertility, age at first birth and AF in the population-based cohort from Norway (the HUNT study) among women over 45 years old. Our findings reveal that women with four or more births, or a history of infertility, or younger age at first birth have approximately a 20% higher risk of AF.

Association of age at menarche, reproductive lifespan and age at menopause with the risk of atrial fibrillation: The HUNT study.

BACKGROUND: Age at menarche, reproductive lifespan, and age at menopause are associated with several cardiovascular diseases, but their relationship with atrial fibrillation (AF) is uncertain. METHODS: We linked information on all women who participated in the third survey of the population-based, longitudinal HUNT study in Norway with medical records from all local hospitals. A total of 14,632 women aged 60 or more were followed for validated incident AF. We retrieved age at menarche and age at menopause from the HUNT questionnaires. Reproductive lifespan was defined as the difference between age at menarche and age at menopause. We used Cox proportional hazards regression models to assess associations between AF and age at menarche, reproductive lifespan, and age at menopause. RESULTS: During a median follow-up of 8.17 years (136,494 person-years), 1217 (8.3 %) participants developed AF. In multivariable-adjusted analyses, we observed no associations between early or late age at menarche and AF (hazard ratios (HRs): <12 years: 0.85 [95 % confidence interval (CI), 0.65-1.12]; ≥16 years: 0.99 [95 % CI, 0.80-1.24] compared to those who attained menarche at 13-14 years). The HR for a reproductive lifespan shorter than 30 years was 0.91 [95 % CI, 0.72-1.15] compared to 34-37 years. Likewise, there was no clear association between premature or early age at menopause and AF (HRs: <40 years: 1.21 [95 % CI, 0.83-1.75]; 40-44 years: 0.97 [95 % CI, 0.77-1.22] compared to 50-54 years). CONCLUSIONS: In this population of women aged 60 years and over, the risk of AF was not associated with age at menarche, reproductive lifespan, or age at menopause.

Clustering phosphate and iron-related markers and prognosis in dialysis patients.

BACKGROUND: Hyperphosphatemia in patients undergoing dialysis is common and is associated with mortality. Recently, the link between phosphate metabolism and iron dynamics has received increasing attention. However, the association between this relationship and prognosis remains largely unexplored. METHODS: We conducted an observational study of patients who initiated dialysis in the 17 centers participating in the Aichi Cohort Study of the Prognosis in Patients Newly Initiated into Dialysis. Data were available on sex, age, use of phosphate binder, drug history, medical history and laboratory data. After excluding patients with missing values of phosphate, hemoglobin, ferritin and transferrin saturation, we used the Gaussian mixture model to divide the cohort into clusters based on phosphate, hemoglobin, logarithmic ferritin and transferrin saturation. We investigated the prognosis of patients in these clusters. The primary outcome was all-cause death. In each cluster, the prognostic impact of phosphate binder was also studied. RESULTS: The study included 1175 patients with chronic kidney disease who initiated dialysis between October 2011 and September 2013. Among them, 785 were men and 390 were women, with a mean ± SD age of 67.9 ± 13.0 years. The patients were divided into three clusters, and mortality was higher in cluster c than in cluster a (P = 0.005). Moreover, the use of phosphate binders was associated with a lower risk of all-cause death in two clusters (a and c) that were characterized by older age and higher prevalence of diabetes mellitus, among other things. CONCLUSIONS: We used an unsupervised machine learning method to cluster patients, using phosphate, hemoglobin and iron-related markers. In two of the clusters, the oral use of a phosphate binder might improve prognosis.

Peripheral artery disease at the time of dialysis initiation and mortality: a prospective observational multicenter study.

OBJECTIVES: Patients with peripheral artery disease (PAD) are reported to have a poorer prognosis than those without PAD. PAD is sometimes found at dialysis initiation, but its influence on the prognosis in these patients has not been investigated. We aimed to compare the mortality rate between patients with PAD at the time of dialysis initiation and those without PAD. DESIGN: We undertook an observational prospective multicenter study of patients starting dialysis treatment. Data were collected on patients' sex, age, presence of PAD, medication, medical history and clinical and laboratory data. SETTING: Seventeen centers participated in the Aichi Cohort Study of Prognosis in Patients Newly Initiated into Dialysis. PARTICIPANTS: A total of 1524 patients with chronic kidney disease started dialysis from October 2011 to September 2013. The patients were followed-up until March 2015. During this time, there were two patients who lost the follow-up. PRIMARY AND SECONDARY OUTCOME MEASURES: The primary outcome was defined as all-cause mortality. The secondary outcomes were defined as each cause of mortality. RESULTS: This study included 1030 men and 492 women with a mean age of 67.50±13.10 years. Of these, 71 had PAD and 1451 did not have PAD. After a median follow-up of 814.5 days, 33.80% of the former group and 17.00% of the latter group had died in March 2015 (p=0.001). After adjusting for confounding factors, PAD at dialysis initiation remained an independent risk factor for mortality (p<0.01). CONCLUSIONS: Patients with PAD at the time of dialysis initiation had a poorer prognosis than patients without PAD. Therefore, the presence of PAD in patients starting dialysis should be considered for their monitoring and follow-up.

Prognostic Impact of Parameters of Metabolic Acidosis in Critically Ill Children with Acute Kidney Injury: A Retrospective Observational Analysis Using the PIC Database.

Acute kidney injury (AKI) is a major complication of sepsis that induces acid-base imbalances. While creatinine levels are the only indicator for assessing the prognosis of AKI, prognostic importance of metabolic acidosis is unknown. We conducted a retrospective observational study by analyzing a large China-based pediatric critical care database from 2010 to 2018. Participants were critically ill children with AKI admitted to intensive care units (ICUs). The study included 1505 children admitted to ICUs with AKI, including 827 males and 678 females. The median age at ICU admission was 22 months (interquartile range 7-65). After a median follow-up of 10.87 days, 4.3% (65 patients) died. After adjusting for confounding factors, hyperlactatemia, low pH, and low bicarbonate levels were independently associated with 28-day mortality (respective odds ratio: 3.06, 2.77, 2.09; p values: <0.01, <0.01, <0.01). The infection had no interaction with the three parameters. The AKI stage negatively interacted with bicarbonate and pH but not lactate. The current study shows that among children with AKI, hyperlactatemia, low pH, and hypobicarbonatemia are associated with 28-day mortality.

The albumin-bilirubin score as a predictor of outcomes in Japanese patients with PBC: an analysis using time-dependent ROC.

The albumin-bilirubin (ALBI) score is calculated using only serum albumin and bilirubin levels, and was developed as a simple method to assess hepatic function. In this study, a total of 409 patients with primary biliary cholangitis (PBC) were enrolled between March 1990 and October 2018. The predictive performances of the ALBI score and other well-established prognostic scores were compared using time-dependent receiver operating characteristic (ROC) analysis. During the follow-up period, 60 patients died, 45 due to liver-related diseases and 15 due to non-liver-related diseases, and 16 patients underwent liver transplantation. Time-dependent ROC analysis showed that the ALBI score has higher the areas under the ROC curves (AUROCs) than the Child-Pugh (C-P) score at each time point; AUROCs at 3, 5, and 10 years after the start of follow-up were 0.94, 0.91, and 0.90 for the ALBI score, and 0.89, 0.88, and 0.82 for the C-P score, respectively. The ALBI score showed the highest AUROCs within 2 years after the start of observation; beyond 2 years, however, the Mayo score had better prognostic ability for mortality and liver transplantation. The ALBI score/grade, derived from objective blood tests, and the Mayo score were superior prognostic tools in PBC patients.

Advanced dynamic statistical parametric mapping with MEG in localizing epileptogenicity of the bottom of sulcus dysplasia.

OBJECTIVE: To investigate whether advanced dynamic statistical parametric mapping (AdSPM) using magnetoencephalography (MEG) can better localize focal cortical dysplasia at bottom of sulcus (FCDB). METHODS: We analyzed 15 children with diagnosis of FCDB in surgical specimen and 3 T MRI by using MEG. Using AdSPM, we analyzed a ±50 ms epoch relative to each single moving dipole (SMD) and applied summation technique to estimate the source activity. The most active area in AdSPM was defined as the location of AdSPM spike source. We compared spatial congruence between MRI-visible FCDB and (1) dipole cluster in SMD method; and (2) AdSPM spike source. RESULTS: AdSPM localized FCDB in 12 (80%) of 15 children whereas dipole cluster localized six (40%). AdSPM spike source was concordant within seizure onset zone in nine (82%) of 11 children with intracranial video EEG. Eleven children with resective surgery achieved seizure freedom with follow-up period of 1.9 ±â€¯1.5 years. Ten (91%) of them had an AdSPM spike source in the resection area. CONCLUSION: AdSPM can noninvasively and neurophysiologically localize epileptogenic FCDB, whether it overlaps with the dipole cluster or not. SIGNIFICANCE: This is the first study to localize epileptogenic FCDB using MEG.