South AfricAn renAl regiStry Annual report 2015

The fourth annual report of the South African Renal Registry summarises the 2015 data on renal replacement therapy (RRT) for patients with end-stage renal disease (ESRD) in South Africa. The South African population increased to 54.96 million in 2015, from 54.00 million in 2014. In December 2015, the number of patients with ESRD who were treated with chronic dialysis or transplantation stood at 10 360, a prevalence of 189 per million population (pmp). The prevalence was 167 pmp in 2013 and 178 pmp in 2014. The increasing prevalence observed is due mainly to the increased numbers of patients accessing haemodialysis in the private sector. In the public sector, which serves 84% of the South African population, the prevalence of RRT (71.9 pmp in 2015) remains at levels close to those reported in 1994 so that the disparity in access continues to increase. The disparities between provinces remain, with Limpopo and Mpumalanga the most under-served, as do the disparities between ethnic groups, with Blacks being the most under-served group.


INTRODUCTION
The South African Renal Registry (SARR) collects, analyses and publishes information on the treatment of patients with kidney failure in South Africa on behalf of the South African Nephrology Society. This is the ninth consecutive annual report published by the SARR, which summarises the data on record for December 2020 on kidney replacement therapy (KRT) for patients with kidney failure in South Africa.
The COVID-19 pandemic resulted in many deaths of our patients on KRT. Studies from South Africa [1] and elsewhere [2] reported mortality rates from COVID-19 approximating 20% in patients on maintenance dialysis, and exceeding 20% in kidney transplant recipients. Transplantation programmes were suspended because of concerns regarding the consequences of SARS-CoV-2 infection in immunocompromised patients, and the management of potential deceased donors was stopped so that emergency departments and intensive care units could prioritise patients with COVID-19 [3]. In the public healthcare sector, where opening new slots for dialysis is largely dependent on an active transplant programme, this led to a further decrease in the access to KRT for patients with kidney failure.
The pandemic also presented major obstacles to the usual operations of the registry. Data capturers had difficulty accessing treatment centres during the lockdown periods, the staff at these facilities were overwhelmed with the challenges of managing cases of COVID-19 and many contracted the disease themselves.

Registry platform
Our current platform was developed using the Webdev programming environment (www.windev.com) and resides on a secure, dedicated, Windows 10 server at a South African internet hosting company. It runs Windows Internet Information Services (IIS) and uses the client/ server version of HFSQL (formerly Hyperfile SQL) as its relational database management system. Data capturers interface with the central database via user-friendly web pages from any device with internet access. The platform uses end-to-end encryption and full backups are made daily.
The quality of our data has improved considerably since we began cross-checking the identity numbers of our patients with the Department of Home Affairs database of births and deaths, which is accessible via the South African Medical Research Council. This has allowed us to analyse and report on patient survival [4] and, more recently, on the survival of elderly patients starting KRT [5].
Over the past few years, the technology platform of the SARR has been expanded to serve as the backbone of the African Renal Registry. Botswana, Burundi, Ghana, Kenya, Nigeria and Zambia have joined the African Renal Registry and have commenced data collection with the aid of our platform [6].

Definitions
Kidney failure and start date of KRT. Kidney failure refers to advanced, irreversible kidney disease which requires the initiation of KRT. The start date is the date of first haemodialysis (HD), the date of the first peritoneal dialysis (PD) flushes or exchanges, or the date of preemptive transplantation (where there is no prior dialysis). For patients who are initially thought to have acute kidney injury (AKI) and are dialysed but who do not recover function and then continue KRT, the start date is the date of the first dialysis, even though the diagnosis at that time was AKI and not kidney failure.
Initial KRT modality. This is the intended first modality and should normally be the modality being used on day 91 of KRT. This means that someone who presents late and who is started on urgent HD but is soon established on PD, will have PD recorded as the initial modality.
Changes in the responsible treating unit. This refers to a change in the dialysis unit, PD follow-up unit/clinic or transplant follow-up unit/centre/practice. A transfer entry in the registry is required to record this. This is not done for short-term transfers when the intention is that the patient will return to the "home" unit, for example, for holiday dialysis, temporary transfer to a unit with isolation facilities, etc.
Primary kidney disease. Responsible nephrologists/ physicians should assist their data-capturers to ensure that this critical information is accurate. We are using the diagnostic codes of the ERA registry [7]. If there is uncertainty about the diagnosis, as is often the case with patients who present late, then it should be recorded as "chronic kidney disease (CKD) -aetiology uncertain/ unknown". In patients who present with kidney failure, small kidneys and hypertension, there should not be an automatic default to labelling such patients as having "chronic glomerulonephritis" or "hypertensive kidney disease".
Chronic hypertensive nephropathy or malignant hypertensive nephropathy. This should be selected as the primary kidney disease only if there is no reason to suspect that the hypertension is secondary to pre-existing renal disease. We suggest that the following criteria be met: hypertension known to precede kidney dysfunction, left ventricular hypertrophy, proteinuria <2 g/day and no evidence of other kidney diseases [8,9].
Lost to follow-up. The SARR assumes that a functioning transplant is maintained unless there is evidence of a "transplant failure" or death. A dialysis modality is assumed to continue for one year from the date of the last registry entry or laboratory result, in the absence of evidence of death; thereafter, the patient is considered lost to followup. Patients are also considered lost to follow-up one year after a "transplant failure" entry if no further entries are recorded.

Recovered kidney function.
Patients who have been initiated on chronic HD/PD and who no longer require dialysis are removed from the registry. The period of dialysis-free recovery must persist for at least 90 days; if the period of recovery is less than 90 days and dialysis is restarted, there should be no END entry and dialysis is considered to have been continuous. If the period of recovery exceeds 90 days and the patient restarts KRT, a new entry is recorded for the patient.

Ethical approval
The SARR operates as a longitudinal study with ethical approval from the Health Research Ethics Committee of Stellenbosch University (reference no. N11/01/028). This is renewed annually upon submission of a progress report. A waiver of individual informed consent has been granted, and the approval includes countrywide data collection on adults and children, in the public and private sectors, and the tapping of various data sources to improve the accuracy and completeness of data. These include records available through doctors' practices, dialysis and transplant centres, provider companies and medical aid funds. Ethical approval has also been granted for the use of the expanded SARR platform for the African Renal Registry. Figure 1 illustrates the provinces and major cities of South Africa. According to the Statistics South Africa (Stats SA) mid-year estimates for 2020 [10], the population of South Africa had increased to 59.62 million people. There was a slight female predominance (51.1%) and Black/African citizens constituted 80.8% of the population (Table 1). About 28.6% of the population was younger than 15 years of age and approximately 9.1% was 60 years or older. The province of Gauteng was home to 26.0% of the population, followed by KwaZulu-Natal with 19.3% (Table 2). Within South Africa, migration has a major impact on the age structure and distribution of provincial populations. For the period 2016-2021, Gauteng and the Western Cape experienced the largest net inflows of migrants, estimated at 980 398 and 290 555, respectively [10].

South Africa in 2020
South Africa is classified as an upper-middle-income country by the World Bank, with a gross national income per capita for 2020 by the Atlas method (current US$) of $6 010 and by the purchasing power parity (PPP) method (current international US$) of $13 140 [11]. Most of the The overall HIV prevalence was 13.0%, and 18.7% for adults aged 15-49 [10].
population (85.1%) rely on the public healthcare sector for services, with only a small proportion (14.9%) having medical insurance and accessing private sector health care [12].
Life expectancy at birth for 2020 was estimated at 62.5 years for males and 68.5 years for females. The infant mortality rate was estimated at 23.6 per 1 000 live births.

Treatment centres for dialysis and transplantation
The number of centres contributing data was 296; of these, 262 (88.5%) are privately owned (Table 3 and Appendix  1). Several provinces have increased access for their public sector patients by utilising spare capacity at private haemodialysis centres on a fee-per-treatment basis. There are also a few privately run centres on the premises of government hospitals which serve public sector patients.

Prevalence and incidence of renal replacement therapy
The total number of patients on KRT on 31 December 2020 was 8 734. This is a prevalence of 146 per million population (pmp). The province with the highest patient numbers remained Gauteng, followed by the Western Cape and KwaZulu-Natal, whereas the province with the highest prevalence was the Western Cape, followed by the Free State and Gauteng ( Figure 2).
There were 672 patients who started KRT in 2020, an incidence of 11.3 pmp. Most of these patients (73.8%) were treated in private centres.  The number of patients treated in the public sector declined, with a prevalence of 44 pmp (Table 4). In the private sector, the prevalence was 729 pmp. The numbers of patients and prevalences by province and healthcare sector are shown in Table 5 and Figure 3. Denominators for prevalence calculations are based on the Stats SA midterm estimates [10] and the Council for Medical Schemes Annual Report [12]. Medical aid beneficiaries who were unclassified with respect to province were allocated to provinces in proportion to the numbers of beneficiaries in each province.

Treatment modality and KRT vintage
Of the patients on KRT in December 2020, 20.7% had a functioning kidney transplant. Of the patients on dialysis, 88.5% were on haemodialysis and 11.5% were on peritoneal dialysis. Most of the transplant and peritoneal dialysis patients were in the public sector; the private sector had much lower proportions of patients on these KRT modalities (Figures 4 and 5).
Overall, the median KRT vintage was 5.

Demographic and clinical data
The median age of the patients on KRT was 53.0 years (IQR 42.4-62.5 years) and 59.8% were male. Because of the rationing and selection criteria applied in public sector hospitals, patients treated there were much younger than those treated in the private sector (44.3 versus 56.0 years). Just more than half of the patients were Black. However, the prevalence was still lowest in Blacks (95 pmp) and highest in Indians/Asians (661 pmp) ( Figure 6).

161
The most common primary kidney disease was hypertensive kidney disease, followed by CKD/kidney failure of unknown cause and diabetic nephropathy (Table 6).
Of the patients with data on diabetes status (8 042 patients), 38.0% had diabetes, with a much higher percentage in the private than in the public sector (45.5% versus 16.6%). The seropositive rate for hepatitis B virus was 2.3% (173 of 7 629 patients), for hepatitis C virus 0.5% (32 of 7 059 patients) and for HIV 12.1% (881 of 7 309 patients).

DISCUSSION
The number of patients on KRT in South Africa stood at 8 734 in December 2020, a prevalence of 146 pmp. These numbers are substantially lower than those reported for December 2019, when the total number of patients treated was 9 937 and the prevalence 169 pmp [13]. We speculate that this is the result of several factors, including larger numbers of deaths (from COVID-19 and other causes), delayed initiation of KRT due to the pandemic, and challenges with data submission to the registry during a period when personnel were overwhelmed with clinical responsibilities related to the pandemic.
The data presented in this report must therefore be interpreted with caution. The many missing year-end entries led to more patients being classified as "lost to follow-up". Some of these patients may still be alive and receiving KRT. In addition, it is likely that some new patients who had started KRT had not yet been entered into the registry. The impact of these factors is a lower reported prevalence.