Chapter 4 UNDERSTANDING SUPPORT FOR TISSUE DONATION ACROSS
5.2. THE EVOLUTION OF THE LIVING DONOR TRANSPLANTATION
5.2.3. A study of the transplant population in Northern Ireland from 2010 to 2017
178
‘real-world’ data set from Belfast Trust to demonstrate how the profile and survival outcomes of ESRD patients with transplanted kidney in Northern Ireland have changed over recent years. This section also illustrates the survival benefit of LDKT versus DDKT in this population. These analyses complement previous discussions and a discussion of the implications are presented in section 5.2.4.
5.2.3.1. Data source
Data were extracted from a population comprising all individuals aged 16 and above who underwent a kidney transplantation at Belfast Trust from January 2010. This time point was chosen because this year marked the initiation of changes in LDKT practices in Northern Ireland [196]. The transplant cohort comprised all recipients of the first kidney transplantation performed during January 2010 until December 2017.
Recipients were followed up until March 2019. This approach allows at least 1-year follow-up data on the transplant cohort. Belfast Trust provided the nature of transplant (including DDKT – deceased donor kidney transplantation, and LDKT – living donor kidney transplantation).
Belfast Trust provided the patient’s health and care number to the Business Service Organization (BSO) for the studied cohorts (comprising LDKT and DDKT cohorts).
BSO has an established ‘Honest Broker Service’ (HBS) that serves as a trusted third party environment to enable the utilization of non-identifiable data for planning, commissioning of services, public health monitoring, and research to improve quality of care [623]. Regardless of the anonymised nature of data, confidentiality issues must be resolved before the data can be accessed in the HBS. A research proposal was submitted to HBS in which signatures from various parties were gathered, sponsorship was addressed, and two independent reviews of the proposal had to be organised. The proposal included a discussion of the ethical issues of the study which principally concerned donor/recipient anonymity and the safeguards adopted to ensure these could be adequately addressed. That data was pseudo-anonymised and no analysis reported where the cell count was below 10 provided sufficient guarantees that anonymity would be preserved and the study was therefore considered exempt by BSO from formal ethics approval. Data must, however, be accessed in person using the HBS
179
secure location which as noted became an issue with the lockdown in response to the COVID19 pandemic.
5.2.3.2. Statistical analysis
Survival analysis was conducted to compare all-cause mortality of LDKT with DDKT.
Patients were followed (statistically) from the index date (date of transplant) to the date of death or censoring (the end of study period (March 2019) or loss to follow-up).
The primary study outcome was the relative risk of death in recipients of a LDKT compared with DDKT. As it was not possible to access data on treatment initiation for dialysis group survival analysis on this group versus transplantation group was not conducted.
Kaplan-Meier curves were used to determine the absolute patient survival probabilities of the LDKT cohort versus DDKT cohort, and of two cohorts based on year of transplantation. A pragmatic approach was applied to achieve equally sized groups:
people who received kidney transplantation from 2010 to 2013 and from 2014 to 2017.
The equality of the survivor functions across the cohorts was tested using the log-rank test. Multivariable Cox proportional-hazards models were used to assess the mortality risk of LDKT versus DDKT cohort after adjusting for potential confounding factors.
Survival models were adjusted for variables commonly associated with survival among patients undergoing any RRTs that were available in the BSO datasets. Those included recipient age at receiving transplant, recipient gender, diabetes status, and clinical variables (primary renal disease diagnosis of recipient, time on RRT prior to transplant, RRT modality prior to transplant, pre-emptive kidney transplant, LOS in hospital for the transplant as an indicator of complexity, and number of HLA-ABDR mismatched). Donor related variables including donor age and gender were also examined. All covariates with a p-value less than 0.25 in the univariate survival model were included in the full model (model 1). A backward elimination approach was used to gradually find a reduced model (model 2).
180
5.2.3.3. Characteristics of the transplant population in Northern Ireland from 2010 to 2017
The profile of kidney transplant recipients has changed considerably between 2010- 2013 and 2014-2017. It is evident from Figure 5.3 that the increased number of transplantation was due largely to the increase in donors aged over 40. The number of recipients aged over 40 also increased significantly. There was a clear shift of age distribution over time where transplantation for older recipients from older donors became more common.
Figure 5.3. Age distribution of recipients and donors (stratified by time period)