Clinical Trial Details
— Status: Active, not recruiting
Administrative data
| NCT number |
NCT03771703 |
| Other study ID # |
EK Nr. 1363/2016 |
| Secondary ID |
|
| Status |
Active, not recruiting |
| Phase |
|
| First received |
|
| Last updated |
|
| Start date |
December 2, 2018 |
| Est. completion date |
April 1, 2024 |
Study information
| Verified date |
October 2023 |
| Source |
Medical University of Vienna |
| Contact |
n/a |
| Is FDA regulated |
No |
| Health authority |
|
| Study type |
Observational
|
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.
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
