And by the Slovenian Ministry of Education, Science and Sport grant 430-168/2013/91. We thank the colleagues Dalibor Fiala, Ludo Waltman and Nees Jan van Eck for useful comments and discussions.Author ContributionsConceived and designed the experiments: LS MB. Performed the experiments: LS. Analyzed the data: LS BMB AK. Contributed reagents/materials/Luteolin 7-O-��-D-glucoside chemical information analysis tools: AK. Wrote the fpsyg.2017.00209 paper: ZL LS.PLOS ONE | DOI:10.1371/journal.pone.0127390 May 18,13 /Consistency of Databases
Potassium abnormalities are very common in dialysis patients, and both hyperkalemia and hypokalemia have been consistently associated with a high risk of all-cause and cardiovascular mortality [1]. Hypokalemia is found in approximately 35 of peritoneal dialysis (PD) patients [2?]. The contribution of hypokalemia to the risk of mortality in PD patients is considerably higher than the one observed in hyperkalemia [5]. A large cohort study performed by Torl et al., analyzing data of more than 120,000 dialysis patients (of which approximately 10,000 were on PD), observed that the populationattributable risk for all-cause mortality was 3.6 for hypokalemia and 1.9 for hyperkalemia [1]. This study raised some important and unanswered questions: first, does potassium fluctuation affect clinical outcomes? And second, is there a causal relationship or hypokalemia is merely a surrogate marker of malnutrition and other comorbidities? Xu et al addressed the first question [6], in an analysis of 886 incident PD patients, demonstrating the effect of potassium variability (expressed as the within-patient standard deviation) on all-cause mortality. The authors concluded that higher serum potassium variability was associated with an independent increase in mortality risk, with the higher quartile presenting an adjusted hazard ratio of 2.43 (CI95 1.03?.46). Nevertheless, the second question remains unanswered, since randomized clinical trials analyzing the causal relationship between hypokalemia and mortality would not be feasible due ethical reasons, and observational studies are associated with selection bias. One interesting approach to minimize the differences between groups would be a propensity match score. When used properly, the propensity match score improves considerably the HMR-1275 chemical information balance of main characteristics between groups. Therefore, the aim of this study was to compare hypokalemic patients with patients with normal serum potassium levels in relation to all-cause, infectious and cardiovascular mortality in large Brazilian 1.07839E+15 PD cohort using the propensity match score.Population and MethodsThis is a nationwide prospective cohort study (BRAZPD II), launched in December 2004, which followed patients until January 2011, described in detail in previous publications [7]. The administrative structure of the BRAZPD II comprises a steering committee with 3 members, one project manager, one project coordinator and one biostatistician. The ethical committees of all participating centers approved the study. The list of all ethic review boards that approved the study can be found in (S1 File). All patients provided written consent, which was approved by the ethical committee and stored locally only in Portuguese. The database contains data from 122 dialysis centers of all regions from Brazil. Any person can submit a project to use data from BRAZPD II for analysis, under the supervision of the Steering Committee. The number of prevalent patients in each year corresponded to 65 to 70 of all P.And by the Slovenian Ministry of Education, Science and Sport grant 430-168/2013/91. We thank the colleagues Dalibor Fiala, Ludo Waltman and Nees Jan van Eck for useful comments and discussions.Author ContributionsConceived and designed the experiments: LS MB. Performed the experiments: LS. Analyzed the data: LS BMB AK. Contributed reagents/materials/analysis tools: AK. Wrote the fpsyg.2017.00209 paper: ZL LS.PLOS ONE | DOI:10.1371/journal.pone.0127390 May 18,13 /Consistency of Databases
Potassium abnormalities are very common in dialysis patients, and both hyperkalemia and hypokalemia have been consistently associated with a high risk of all-cause and cardiovascular mortality [1]. Hypokalemia is found in approximately 35 of peritoneal dialysis (PD) patients [2?]. The contribution of hypokalemia to the risk of mortality in PD patients is considerably higher than the one observed in hyperkalemia [5]. A large cohort study performed by Torl et al., analyzing data of more than 120,000 dialysis patients (of which approximately 10,000 were on PD), observed that the populationattributable risk for all-cause mortality was 3.6 for hypokalemia and 1.9 for hyperkalemia [1]. This study raised some important and unanswered questions: first, does potassium fluctuation affect clinical outcomes? And second, is there a causal relationship or hypokalemia is merely a surrogate marker of malnutrition and other comorbidities? Xu et al addressed the first question [6], in an analysis of 886 incident PD patients, demonstrating the effect of potassium variability (expressed as the within-patient standard deviation) on all-cause mortality. The authors concluded that higher serum potassium variability was associated with an independent increase in mortality risk, with the higher quartile presenting an adjusted hazard ratio of 2.43 (CI95 1.03?.46). Nevertheless, the second question remains unanswered, since randomized clinical trials analyzing the causal relationship between hypokalemia and mortality would not be feasible due ethical reasons, and observational studies are associated with selection bias. One interesting approach to minimize the differences between groups would be a propensity match score. When used properly, the propensity match score improves considerably the balance of main characteristics between groups. Therefore, the aim of this study was to compare hypokalemic patients with patients with normal serum potassium levels in relation to all-cause, infectious and cardiovascular mortality in large Brazilian 1.07839E+15 PD cohort using the propensity match score.Population and MethodsThis is a nationwide prospective cohort study (BRAZPD II), launched in December 2004, which followed patients until January 2011, described in detail in previous publications [7]. The administrative structure of the BRAZPD II comprises a steering committee with 3 members, one project manager, one project coordinator and one biostatistician. The ethical committees of all participating centers approved the study. The list of all ethic review boards that approved the study can be found in (S1 File). All patients provided written consent, which was approved by the ethical committee and stored locally only in Portuguese. The database contains data from 122 dialysis centers of all regions from Brazil. Any person can submit a project to use data from BRAZPD II for analysis, under the supervision of the Steering Committee. The number of prevalent patients in each year corresponded to 65 to 70 of all P.
