Mortality prediction with adaptive feature importance recalibration for peritoneal dialysis patients.
Ontology highlight
ABSTRACT:
SUBMITTER: Ma L
PROVIDER: S-EPMC10724364 | biostudies-literature | 2023 Dec
REPOSITORIES: biostudies-literature
Publications
Mortality prediction with adaptive feature importance recalibration for peritoneal dialysis patients.
Ma Liantao L Zhang Chaohe C Gao Junyi J Jiao Xianfeng X Yu Zhihao Z Zhu Yinghao Y Wang Tianlong T Ma Xinyu X Wang Yasha Y Tang Wen W Zhao Xinju X Ruan Wenjie W Wang Tao T
Patterns (New York, N.Y.) 20231208 12
The study aims to develop AICare, an interpretable mortality prediction model, using electronic medical records (EMR) from follow-up visits for end-stage renal disease (ESRD) patients. AICare includes a multichannel feature extraction module and an adaptive feature importance recalibration module. It integrates dynamic records and static features to perform personalized health context representation learning. The dataset encompasses 13,091 visits and demographic data of 656 peritoneal dialysis ( ...[more]
Similar Datasets
Project description:BackgroundResults concerning the association between peritoneal dialysis-related peritonitis and mortality in peritoneal dialysis patients are inconclusive, with one potential reason being that the time-dependent effect of peritonitis has rarely been considered in previous studies. This study aimed to evaluate whether peritonitis has a negative impact on mortality in a large cohort of peritoneal dialysis patients. We also assessed the changing impact of peritonitis on patient mortality with respect to duration of follow-up.MethodsThis retrospective cohort study included incident patients who started peritoneal dialysis from 1 January 2006 to 31 December 2011. Episodes of peritonitis were recorded at the time of onset, and peritonitis was parameterized as a time-dependent variable for analysis. We used the Cox regression model to assess whether peritonitis has a negative impact on mortality.ResultsA total of 1321 patients were included. The mean age was 48.1 ± 15.3 years, 41.3% were female, and 23.5% with diabetes mellitus. The median (interquartile) follow-up time was 34 (21-48) months. After adjusting for confounders, peritonitis was independently associated with 95% increased risk of all-cause mortality (hazard ratio, 1.95; 95% confidence interval: 1.46-2.60), 90% increased risk of cardiovascular mortality (hazard ratio, 1.90; 95% confidence interval: 1.28-2.81) and near 4-fold increased risk of infection-related mortality (hazard ratio, 4.94; 95% confidence interval: 2.47-9.86). Further analyses showed that peritonitis was not significantly associated with mortality within 2 years of peritoneal dialysis initiation, but strongly influenced mortality in patients dialysed longer than 2 years.ConclusionsPeritonitis was independently associated with higher risk of all-cause, cardiovascular and infection-related mortality in peritoneal dialysis patients, and its impact on mortality was more significant in patients with longer peritoneal dialysis duration.
| S-EPMC5460447 | biostudies-literature
Project description:Background New lipid-lowering therapy at the start of dialysis and measurement of lipid parameters over the follow-up period is not recommended in dialysis patients, which seems unappropriated in clinical practice. We aimed to examine the effect of hyperlipidemia on mortality in patients undergoing continuous ambulatory peritoneal dialysis (CAPD). Methods A retrospective cohort study was performed, including 2939 incident CAPD patients from five dialysis facilities between January 1, 2005, and December 31, 2018. The primary outcome was all-cause mortality. The association between hyperlipidemia at the start of CAPD and all-cause mortality was evaluated using Cox proportional hazards regression. Results Of 2939 with a median age of 50.0 (interquartile range, 39.0–61.0), 1697 (57.7%) were men, 533 (18.1%) had hyperlipidemia, 549 (18.7%) had diabetes mellitus, 1915 (65.2%) had hypertension, and 410 (14.0%) had a history of CVD. During the median follow-up period of 35.1 months, 519 (17.7%) died, including 402 (16.7%, 47.4/1000 patient-years) in the non-hyperlipidemia group and 117 (22.0%, 71.1/1000 patient-years) in the hyperlipidemia group. Over the overall follow-up period, patients with hyperlipidemia had an equally high risk of all-cause mortality throughout follow-up as those without hyperlipidemia ([HR] 1.04, 95% confidence interval [CI] 0.83 to 1.31). However, from the 48-month follow-up onwards, hyperlipidemia was associated with a 2.26 (95% CI 1.49 to 3.43)-time higher risk of all-cause mortality than non-hyperlipidemia. Hypertension modified the association between hyperlipidemia and all-cause mortality (P for interaction < 0.001). A significantly increased risk of all-cause mortality was observed among patients with hypertension (HR 2.27, 95%CI 1.44–3.58). Conclusion Among CAPD patients, hyperlipidemia at the beginning of CAPD was associated with a high risk of long-term mortality. Hypertension may mediate the association. Our findings suggested that long-term lipid-lowering treatment should be used in those patients with hyperlipidemia. Supplementary information The online version contains supplementary material available at 10.1186/s12882-022-02970-w.
| S-EPMC9590170 | biostudies-literature
Project description:BackgroundAssessment of infection-related mortality remains inadequate in patients undergoing peritoneal dialysis. This study was performed to develop a risk model for predicting the 2-year infection-related mortality risk in patients undergoing peritoneal dialysis.MethodsThe study cohort comprised 606 patients who started and continued peritoneal dialysis for 90 at least days and was drawn from the Fukuoka Peritoneal Dialysis Database Registry Study in Japan. The patients were registered from 1 January 2006 to 31 December 2016 and followed up until 31 December 2017. To generate a prediction rule, the score for each variable was weighted by the regression coefficients calculated using a Cox proportional hazard model adjusted by risk factors for infection-related mortality, including patient characteristics, comorbidities, and laboratory data.ResultsDuring the follow-up period (median, 2.2 years), 138 patients died; 58 of them of infectious disease. The final model for infection-related mortality comprises six factors: age, sex, serum albumin, serum creatinine, total cholesterol, and weekly renal Kt/V. The incidence of infection-related mortality increased linearly with increasing total risk score (P for trend <0.001). Furthermore, the prediction model showed adequate discrimination (c-statistic = 0.79 [0.72-0.86]) and calibration (Hosmer-Lemeshow test, P = 0.47).ConclusionIn this study, we developed a new model using clinical measures for predicting infection-related mortality in patients undergoing peritoneal dialysis.
| S-EPMC6426225 | biostudies-literature
Project description:Premature all-cause mortality is high in patients receiving peritoneal dialysis (PD). The accurate and early prediction of mortality is critical and difficult. Three prediction models, the logistic regression (LR) model, artificial neural network (ANN) classic model and a new structured ANN model (ANN mixed model), were constructed and evaluated using a receiver operating characteristic (ROC) curve analysis. The permutation feature importance was used to interpret the important features in the ANN models. Eight hundred fifty-nine patients were enrolled in the study. The LR model performed slightly better than the other two ANN models on the test dataset; however, in the total dataset, the ANN models fit much better. The ANN mixed model showed the best prediction performance, with area under the ROC curves (AUROCs) of 0.8 and 0.79 for the 6-month and 12-month datasets. Our study showed that age, diastolic blood pressure (DBP), and low-density lipoprotein cholesterol (LDL-c) levels were common risk factors for premature mortality in patients receiving PD. Our ANN mixed model had incomparable advantages in fitting the overall data characteristics, and age is a steady risk factor for premature mortality in patients undergoing PD. Otherwise, DBP and LDL-c levels should receive more attention for all-cause mortality during follow-up.
| S-EPMC8202888 | biostudies-literature
Project description:ObjectiveHypomagnesemia has been associated with an increase in mortality among the general population as well as patients with chronic kidney disease or those on hemodialysis. However, this association has not been thoroughly studied in patients undergoing peritoneal dialysis. The aim of this study was to evaluate the association between serum magnesium concentrations and all-cause and cardiovascular mortalities in peritoneal dialysis patients.MethodsThis single-center retrospective study included 253 incident peritoneal dialysis patients enrolled between July 1, 2005 and December 31, 2014 and followed to June 30, 2015. Patient's demographic characteristics as well as clinical and laboratory measurements were collected.ResultsOf 253 patients evaluated, 36 patients (14.2%) suffered from hypomagnesemia. During a median follow-up of 29 months (range: 4-120 months), 60 patients (23.7%) died, and 35 (58.3%) of these deaths were attributed to cardiovascular causes. Low serum magnesium was positively associated with peritoneal dialysis duration (r = 0.303, p < 0.001) as well as serum concentrations of albumin (r = 0.220, p < 0.001), triglycerides (r = 0.160, p = 0.011), potassium (r = 0.156, p = 0.013), calcium(r = 0.299, p < 0.001)and phosphate (r = 0.191, p = 0.002). Patients in the hypomagnesemia group had a lower survival rate than those in the normal magnesium groups (p < 0.001). In a multivariate Cox proportional hazards regression analysis, serum magnesium was an independent negative predictor of all-cause mortality (hazard ratio [HR] = 0.075, p = 0.011) and cardiovascular mortality (HR = 0.003, p < 0.001), especially in female patients. However, in univariate and multivariate Cox analysis, △Mg(difference between 1-year magnesium and baseline magnesium) was not an independent predictor of all-cause mortality and cardiovascular mortality.ConclusionHypomagnesemia was common among peritoneal dialysis patients and was independently associated with all-cause mortality and cardiovascular mortality.
| S-EPMC4811555 | biostudies-literature
Project description:ObjectiveThis study aimed to evaluate inflammatory biomarkers in patients undergoing peritoneal dialysis and investigate their association with all-cause mortality or transfer to hemodialysis.MethodsThis prospective cohort study included 43 patients undergoing peritoneal dialysis. Plasma levels of cytokines were measured using flow cytometry and capture enzyme-linked immunosorbent assay. Biomarkers were categorized based on their respective median values. Survival analysis was conducted using the Kaplan-Meier estimator, considering two outcomes: all-cause mortality and transfer to hemodialysis.ResultsAfter adjusting for confounding factors, plasma levels above the median of the levels of CCL2 and plasma, as well as below the median of TNF-α, and the median of dialysate IL-17 levels, were associated with an increased risk of experiencing the specified outcomes after approximately 16 months of follow-up.ConclusionThese findings suggest that inflammatory biomarkers may be a valuable tool for predicting all-cause mortality and transfer to hemodialysis in patients undergoing peritoneal dialysis.
| S-EPMC11323835 | biostudies-literature
Project description:BackgroundContinuous ambulatory peritoneal dialysis (CAPD) patients with diabetes are at increased risk of mortality and high peritoneal transporters appear to contribute to poor survival. However, little is known about the combined impacts of high peritoneal transporters and diabetes on mortality.MethodsThis was a prospective observational cohort study. 776 incident CAPD patients were enrolled. Unadjusted and adjusted Cox proportional regression models were used to evaluate the association and interaction of peritoneal transport and diabetic status with mortality.ResultsIn the entire cohort, high peritoneal transport status was associated with an increased risk of all-cause mortality in unadjusted model [hazard ratio (HR) 2.35, 95% confidence interval (CI) 1.30 to 4.25, P = 0.01], but this association was not significant in multivariable model. There was an interaction between peritoneal membrane transport status and diabetes (P = 0.028). Subgroup analyses showed that compared to low and low average transporters, high transporters was associated with a higher risk of all-cause mortality (adjusted HR 1.78, 95% CI 1.07 to 4.70, P = 0.04) in CAPD patients without diabetes, but not in those with diabetes (adjusted HR 0.79, 95%CI 0.33 to 1.89, P = 0.59). Results were similar when transport status was assessed as a continuous variable.ConclusionsThe association between high peritoneal transport and all-cause mortality was likely to vary with diabetes status. High peritoneal transport was associated with an elevated risk of death among CAPD patients without diabetes, but not in those with diabetes.
| S-EPMC4199757 | biostudies-literature
Project description:Objectives:The monocyte-to-lymphocyte ratio (MLR), as a new marker of the systemic inflammatory response, is associated with cardiovascular disease (CVD) mortality in the general population and hemodialysis patients. However, the association between the MLR and CVD mortality in peritoneal dialysis (PD) has received little attention. Methods:In this multicenter retrospective cohort study, 1753 incident PD patients from November 1, 2005, to June 30, 2017, with a baseline MLR were enrolled. The primary endpoint was CVD mortality. The association of MLR with CVD mortality was assessed using a multivariable-adjusted Cox model and the Fine and Gray competing risk model. Results:Of 1753 patients, the mean age was 51.1 ± 14.9 years, 56.9% of patients were male, and the Charlson comorbidity index was 4.29 ± 1.75. During the follow-up period of 31.2 ± 18.4 months, 368 patients died, of which 200 (54.3%) deaths were caused by CVD events. CVD mortality rates for the lowest, middle, and highest MLR tertiles were 70.6, 78.4, and 88.9 per 1000 patient-years, respectively (P < 0.001). Kaplan-Meier analysis revealed that survival rates were significantly different among the three MLR groups (log?rank = 22.41, P < 0.001). Kaplan-Meier analysis revealed that survival rates were significantly different among the three MLR groups (log?rank = 22.41, P < 0.001). Kaplan-Meier analysis revealed that survival rates were significantly different among the three MLR groups (log?rank = 22.41, P < 0.001). Kaplan-Meier analysis revealed that survival rates were significantly different among the three MLR groups (log?rank = 22.41, P < 0.001). After adjusting for confounding factors, the highest MLR tertile was significantly associated with a hazard ratio (HR) for CVD mortality of 1.45 (95% confidence interval, 1.13-2.51, P = 0.016). The Fine and Gray method analysis showed that using all-cause mortality as competing risk, the highest MLR tertile remained an independent predictor of CVD mortality (HR = 1.39, 95% CI 1.10-2.47, P = 0.021). Conclusions:Higher MLR levels at the commencement of PD may be independently associated with increased CVD mortality in PD patients.
| S-EPMC7048939 | biostudies-literature
Project description:BackgroundThe prognostic value of pulmonary hypertension at the start of peritoneal dialysis (PD) in patient survival is unclear.MethodsWe conducted a retrospective study of incident patients who initiated PD therapy from January 2007 to December 2011, and followed up through June 2013. Pulmonary hypertension was defined as an estimated systolic pulmonary artery pressure (PAP) of ≥ 35 mm Hg using echocardiography. Clinical parameters and laboratory findings were compared between patients with and without pulmonary hypertension and a logistic regression model was elaborated. Patient outcomes (all-cause and cardiovascular mortality) were recorded during follow-up. Survival curves were constructed by the Kaplan-Meier method, and the influences of pulmonary hypertension on outcomes were analyzed by Cox regression models.ResultsPulmonary hypertension was prevalent in 99 (16.0%) of the 618 patients studied. The independent risk factors for pulmonary hypertension were female (odds ratio [OR] = 2.12; 95% confidence interval [CI]: 1.29 - 3.46), left atrial diameter (OR = 1.15; 95% CI: 1.10 - 1.20), left ventricular ejection fraction (OR = 0.97; 95% CI: 0.95 - 0.99), and serum sodium (OR = 0.94; 95% CI: 0.89 - 0.99). Over a median follow-up of 29.4 months, 93 patients (15.0%) died, 59.1% of them due to cardiovascular disease. Kaplan-Meier survival analysis showed that patients with pulmonary hypertension had worse overall rates of survival and cardiovascular death-free survival than those without pulmonary hypertension. After multivariate adjustment, pulmonary hypertension was independently associated with increased risk for both all-cause and cardiovascular mortality, with hazard ratios (HRs) of 2.10 (95% CI: 1.35 - 3.27) and 2.60 (95% CI: 1.48 - 4.56), respectively.ConclusionsThe prevalence of pulmonary hypertension at the start of PD was common and associated with increased risk of both all-cause and cardiovascular mortality in incident PD patients.
| S-EPMC4597986 | biostudies-literature
Project description:BackgroundEnd-stage kidney disease (ESKD) treatment is very costly and accounts for a significant percentage of public healthcare expenditures. Beyond direct costs, dialysis patients use other healthcare levels, but the impact of public investment on each of these levels is unclear. This study aimed to investigate the association between direct financing at different healthcare levels and overall mortality in peritoneal dialysis (PD) patients.MethodsWe included all adult incident PD patients from a Brazilian prospective, nationwide PD cohort. Overall mortality was the primary outcome of interest. We used a three-level multilevel survival analysis to investigate whether mortality was associated with the investments destined to different levels of healthcare complexity: (i) primary, (ii) medium and high and (iii) professional healthcare training and community awareness.ResultsWe evaluated 5707 incident PD patients from 78 Brazilian cities, which were divided into four quartiles for each healthcare level (Groups I-IV). After taking the highest quartile (Group IV) as a reference, investment in the primary health level was not associated with patient survival. Otherwise, medium and high complexity levels were associated with higher mortality risk. Also, investment in healthcare manager training and community awareness had an impact on patient survival.ConclusionsInvestments in different levels of the healthcare system have distinct impacts on PD patient survival. Investment in healthcare manager training and community awareness seems to be a promising strategy on which to focus, given the relatively low cost and positive impact on outcome.
| S-EPMC7769512 | biostudies-literature