Sex Differences in Chronic Kidney Disease

Chronic Kidney Diseases Clinical Trial

Official title: An Exploratory Study in Three Parts to Determine Sex-Specific Differences in Chronic Kidney Disease Recognition, Progression and Treatment

Status Active, not recruiting

Clinical Trial Summary

This study aims to describe sex specific differences of the hemodialysis population in Austria, to quantify sex specific differences in treatment and outcomes in hospitalized patients with chronic kidney disease and to examine decision making by doctors and patients with regards to hemodialysis initiation.

To adequately serve the needs of these research questions, the study is divided into 3 parts (sub-studies).

1. Description and analysis of sex-specific differences in renal replacement therapy in Austria.

2. Description and analysis of the competing risks of death versus renal replacement therapy initiation in the chronic kidney disease population by sex.

3. Analysis of sex differences in renal replacement therapy initiation decisions, emphasizing patient perception and socio-economic differences.

Clinical Trial Description

2.1-A Renal Replacement Therapy: Description and analysis of sex-specific differences in hemodialysis patients and the associated male-to-female mortality in Austria

Hypothesis:

There are sex-specific differences in hemodialysis incidence/prevalence and mortality among hemodialysis patients in Austria. There are differences in male-to-female mortality among hemodialysis patients as compared to the general population in Austria.

Methodology:

Data extraction will be performed from the Austrian Dialysis registry which holds complete longitudinal entries of all Austrian hemodialysis patients since 1970 and contains data of all patients undergoing hemodialysis in Austria since the year 1965.Demographic data, laboratory values, clinical characteristics and hemodialysis specific data will be collected for every patient registered in the Austrian Dialysis registry between 1965 and 2015.Male and female population structure and mortality data of the Austrian general population will be retrieved from the ´Statistik Austria´ registry.Additional data of the Austrian general population will be retrieved from the SYSKID project and electronic medical records retrieved in the course of an epidemiologic study.

Statistics:

Standard descriptive statistics will be used for tabulations of patient characteristics by sex and decade, of the Austrian hemodialysis population versus general population by sex and age group over time. The unadjusted male-to-female mortality rate ratio in the hemodialysis population will be compared with that of the Austrian general population. Adjusted hazard ratios for the male-to-female mortality risk in hemodialysis patients, by region will be performed using Cox regression analysis. Statistical modeling will closely follow a previously published study of hemodialysis data, analyzed by sex and country/region. All obtained p-values will be considered exploratory in nature.

2.1-B Renal Replacement Therapy: Description and analysis of sex-specific differences in transplantation rates in Austria

Hypothesis:

(i) Women in Austria have lower access to transplantation with a deceased donor organ.(ii) Living kidney donor transplantation is not unfair towards women. (iii) Kidney allograft survival is sex-dependent.

Methodology:

Demographic and clinical characteristics, laboratory values, and hemodialysis specific data will be extracted from the Austrian Dialysis and Transplant registry for all patients registered between 1965 and 2018. Data regarding the time intervals between initiation of dialysis, entry to transplantation waiting list, transplantation, potential graft failure and potential death will be retrieved. Further, data regarding sex of donors and recipients will be extracted.

Statistics:

Hypothesis (i), sex differences in the time to waiting-list entry and time on waiting list, male-to-female hazard ratios will be calculated from competing risk Cox proportional hazards models. The models will include age and relevant comorbidities as adjusting covariates and death will be considered as competing event to transplantation.

Hypothesis (ii), a logistic regression model will be fit to explain the probability for a dialysis patient to receive a living-donor kidney transplant by sex and a comprehensive set of further potential explanatory variables. These will include age, comorbidities, serological markers, and marital status.

The model will be used to calculate covariate adjusted male-to-female odds ratios. In case of a "fair" treatment of men and women, the observed adjusted effect of sex is expected to be close to zero whereas inter-individual differences in the probability to receive a living-donor transplant should be explained through individual, none sex-related covariate values.

Hypothesis (iii), Cox proportional hazard models will be applied for the competing events graft failure and death, with sex as main predictor and demographic and clinical data on donor and recipient as covariates. In a second model sex differences regarding a combined endpoint comprising graft failure and death will be studied as sensitivity analysis.

2.2 Description and analysis of the competing risks of death versus renal replacement therapy initiation in the chronic kidney disease population by sex

Hypothesis:

There are sex-specific differences in the competing risks of death versus renal replacement therapy initiation in the chronic kidney disease population.

Methodology:

The principal data source from Europe is the Stockholm CREAtinine Measurements (SCREAM) cohort. In the SCREAM cohort, 1 Million adult individuals residing in the region of Stockholm/Sweden and undergoing one or more routine serum creatinine assessment between 2006 and 2011 were included. This number of individuals corresponds to 66% coverage of the complete population in the region. The advantages of using the SCREAM cohort are (a) universal healthcare access of the included individuals, (b) the fact that all laboratory values of the included individuals are provided through the central repository of laboratory data (no other laboratory values than the included ones have ever been taken), (c) complete coverage on ICD diagnoses (including primary care) and death, (d) availability of socio-economic information. The most complex part of SCREAM, linking the laboratory data to regional healthcare utilization data was solved by involving the administrative health data register of the region, entitled Vårdanalysdatabasen, (i.e. the Stockholm regional healthcare data warehouse). As SCREAM remains without loss to follow-up and is likely one of the largest and most complete healthcare utilization cohorts in Europe, it is an ideally suited database to answer the research questions laid out here above. The research results obtained by the US team from analyzing USRDS datasets (NHANES, Optum, CMS) will complement the analyses in SCREAM and will either be published back-to-back, or within the same publication. Additional data will be retrieved from international databases (e.g. NHANES, USRDS).

In this model the hazard ratios for three possible transitions (RRT, disease progression, death) are estimated separately for each CKD stage. The model will further account for the influence of covariates such as comorbidities and socio-economic factors. In case of time-varying covariates the last observed value will be utilized in each estimation step. To support a causal conclusion of sex-related effects, a matched-pairs analysis will be performed. Pairs of one woman and one man will be formed using nearest neighbour matching based on potentially confounding covariates and stratified Cox models, treating each pair as stratum, will be fit to calculate the cause-specific hazard ratios between men and woman for RRT, disease progression and death. Further, sex differences in disease progression will be analyzed by comparing the average rates of eGFR decline with time and the hazard rates for reaching secondary endpoints such as doubling of serum creatinine and a decline in eGFR of more than 30%. The rates of referral to a nephrologist and the frequencies of CKD related diagnoses (defined by appropriate CKD-10 codes) will be analyzed and contrasted between sexes. As an important side-effect, we will attempt to define criteria for the "predialysis stage" outside of acutely life-threatening conditions such as hyperkalemia and pulmonary congestion. After the transition of subjects into our defined "predialysis stage", RRT or death may be the consequence and the competing risks for either, as a function of time, will also be analyzed separately between men and women, and subsequently compared.

2.3 Analysis of sex differences in RRT initiation decisions, emphasizing patient perception and socio-economic differences

Hypothesis:

There are sex-specific differences in decision-making of doctors and patients with end-stage renal disease with regards to renal replacement therapy initiation.

Methodology:

Qualitative study: Criteria for participation in our study will be (i) a diagnosis of CKD (eGFR <60 mL/min/1.73 m2, while 50% of all interviews will be conducted among patients with eGFR <15 mL/min/1.73 m2) and (ii) appropriate language skills. In order to gain a holistic perspective on the topic and in addition to the interviews with the patients, we will invite a family member and/or caregiver per patient to participate in an additional interview. Furthermore, we will interview two health professionals who are clinical experts in CKD from different professions in each center regarding their perspectives on gender differences in CKD and access to dialysis. Data collection will be undertaken through semi-structured interviewing. The first interview will start with an open question regarding the most important area in daily life that the health condition has affected. Patients will be encouraged to consider not only present experiences, but also prior experiences from their life stories. The subsequent interview questions will be based on the World Health Organization International Classification of Functioning, Disability and Health (http://www.who.int/classifications/icf/en/) and the Canadian Occupational Performance Measure (http://www.thecopm.ca/) to cover comprehensively all areas of functioning in daily life. The semi-structure interview guide will be adapted for the family members, care givers and health professionals. Two interviews will be performed with each participant. The purpose of the second interview is to give the researcher and the interviewee time to reflect about what was said in the previous interviews. Interviews will be digitally tape-recorded and transcribed verbatim. The analysis will be supported by using appropriate software, e.g. NVivo or AtlasTi. In order to ensure accuracy and rigor of the qualitative analysis, all interviews will be performed according to a pre-determined standard procedure. 10-15% of the results of the different steps of the analysis will be reviewed by a senior researcher with extensive prior experience in the field of qualitative research and one patient research partner. Triangulation will be achieved by comparison of the results, especially new and unexpected findings to literature, also in other diseases.

Based on prior work, we are estimating that a total number of 1000 patient questionnaires will yield adequate information. This sample size will be modified, depending on the results of the qualitative study. For the longitudinal follow-up, we assume that we will be able to reach 50% of the baseline patients who will agree to fill in the follow-up questionnaire as well. We are planning to include up to 10 representative Nephrology clinics in Austria. Data of the returned questionnaires will either be entered into an IBM SPSS statistics database or analyzed in R (www.r-project.org). The analysis of the survey will include descriptive statistics for all variables which will be depicted according to their distribution, as well as sub-group analyses for gender/sex-differences with interferential statistical tests depending again on the distribution of each variable. Graphical illustrations will be performed, as needed. For multiple testing, Bonferroni corrections will be applied. Based on the concepts identified in the qualitative analysis, their meaning and their relationship to each other, we will select the dependent and the independent variables and fit a regression model to explore the relationship between the concepts/variables of the survey in qualitative terms ;

Related Conditions & MeSH terms

• Chronic Kidney Diseases
• Kidney Diseases
• Renal Insufficiency, Chronic

• NCT number: NCT03771703
• Study type: Observational
• Source: Medical University of Vienna
• Status: Active, not recruiting
• Start date: December 2, 2018
• Completion date: April 1, 2024