Study design and subjects
The DOPPS is an international prospective cohort study of in-center adult HD patients which described in previous published papers [22, 23]. China joined DOPPS in 2011. DOPPS China randomly selected an average of 30 patients from 15 dialysis facilities in each city of Beijing, Shanghai, and Guangzhou. This was described in our previous study [24,25,26]. There were 1427 patients participated in China DOPPS5 (2012–2015). Of the 1427 patients, 58 patients were excluded from the present analysis as they didn’t have the platelet records. Baseline demographic and clinical data were collected at the start of participation in DOPPS5.
The authors confirm that all methods were carried out in accordance with relevant guidelines and regulations.
Participants were divided into 2 groups according to their baseline platelet counts. Patients with thrombocytopenia (platelet< 100*109/L) were assigned as TP group, and patients without thrombocytopenia as Non-TP group (platelet ≥100*109/L). The Non-TP could not be further divided into normal (100*109/L-300*109/L) or above normal (platelet > 300*109/L), since there were few patients who had platelet counts beyond the high-end of normal limits (300*109/L).
The primary end-point event was all-cause mortality. The secondary end-point event was the CV mortality during the follow-up period. We have a ‘Termination form’ to collect patients’ death information, including the date, place, primary reason, secondary reason of death. And the reasons of death were divided into several categories (cardiac, vascular, liver disease, infection, gastrointestinal, metabolic, other).
CV mortality was defined by the primary death records in the dataset. The following diagnosis in primary death records were considered as CV mortality: atherosclerotic heart disease, cardiac arrest, cardiac arrhythmia, cardiomyopathy, cerebro-vascular accident (including intracranial Hemorrhage), congestive heart failure, hemorrhage from ruptured vascular aneurysm, ischemic brain damage/anoxic encephalopathy, acute myocardial infarction, pulmonary embolus, stroke, and valvular heart disease.
Continuous variables were represented as mean ± SD or median (25th, 75th) according to the results of normality test. Categorical variables were expressed as number and percentage. We stratified data by TP and Non-TP groups. Differences in mean or median among groups were evaluated by using analysis of variance or non-parametric test. Categorical data were compared using chi-square test.
Survival curves were produced by the Kaplan-Meier method and estimated by log-rank test. We used Cox proportional hazards models to assess the association of baseline platelet count with all-cause mortality, and CV mortality. All Cox models accounted for facility clustering effects by using the robust sandwich covariance estimate. Survival time for Cox models of all-cause mortality was the time from study entry to the end of study or to death, whichever occurred first. Similar calculation was taken for CV mortality. The Non-TP group was taken as the reference group for all analyses. Cox regression models were with 5 incremental levels of covariate adjustment. Model 1: unadjusted; model 2: adjusted for age, gender, body mass index (BMI), vintage; model 3: model 2 variables plus comorbidities (diabetes, coronary artery disease, congestive heart failure, other cardiovascular disease, cerebrovascular disease, hepatitis B and C, cancer (non-skin), peripheral vascular disease, lung disease, hypertension, psychiatric disorder, GI Bleeding, recurrent cellulitis, fracture, neurologic disease); model 4: model 3 plus hemoglobin, albumin, white blood cells, and serum creatinine; model 5: model 4 plus intradialytic weight loss, fistula use, primary kidney disease, standard kt/v, urine output< 200 ml/day.
We also used stepwise multivariate logistic analysis to identify the impact factors of TP. Odds ratio (OR) and 95% conference interval (CI) were calculated for each variable.
We performed MI procedure to impute missing data, and continuous and categorical variables were imputed 25 times by fully conditional specification regression and logistic regression, respectively. The imputed data sets were analyzed using the MI Analyze procedure in SAS/STAT 9.4. Percentages of missing for most variables were < 10%, except for single-pooled Kt/V (36.2%). P value < 0.05 was considered as statistically significant. All statistical analyses were performed with SAS, version 9.4 (SAS institute, Cary, NC; USA).
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.