End-stage kidney disease and rationing of kidney replacement therapy in the free state province, South Africa: a retrospective study.

Background

Chronic kidney disease (CKD) defined by the persistent kidney damage or glomerular filtration rate (GFR) of < 60 mL/min/1.73 m 2 is a significant contributor to mortality due to non-communicable diseases [ 1 , 2 ]. Its prevalence increases with age, but in Sub-Saharan Africa (SSA), young adults tend to be disproportionately affected [ 3 – 5 ], with higher incidences of chronic glomerulonephritis, herbal/traditional remedies, and communicable diseases in SSA contributing to CKD’s overrepresentation among the younger groups [ 6 , 7 ]. Accurate prevalence statistics in SSA are scarce, but an estimated prevalence rate of 13 % has been reported [ 8 ]. Low-to middle-income countries face many challenges in the prevention and early detection of CKD, leading to an increase in the risk of disease progression [ 8 ]. In the Republic of South Africa (RSA), CKD presents within a context of a rising double burden of communicable and non-communicable diseases [ 9 , 10 ].

End-stage kidney disease (ESKD) is an advanced, irreversible final stage of CKD requiring kidney replacement therapy (KRT) to sustain life [ 1 ]. It is a public health challenge worldwide and poses an enormous economic burden on the healthcare systems [ 11 , 12 ]. With a population of just more than a billion, SSA has 80 % of its population living on less than $2.5 a day [ 13 ]. Access to the full range of KRT options, including dialysis and kidney transplantation, is a challenge for most SSA countries due to the high costs, infrastructural constraints, and shortage of skilled personnel [ 14 ]. Estimates suggest a significant unmet need in Africa, projected to have the lowest KRT access rate in the world, ranging between 9 and 16 % [ 14 ]. Furthermore, most dialysis centres are located in urban areas creating internal disparities in access to a full range of ESKD management options [ 15 ]. This situation favours access to KRT by those in urban areas with the ability to fund treatment costs.

In RSA, those who can afford private insurance, approximately 16 % of the population, receive a full range of KRT modalities. At the same time, the majority who are mostly indigent rely on the public sector [ 16 ]. Thus, the RSA government adopted a KRT rationing policy for the public sector based on criteria that select patients considered to be the best candidates for transplantation to balance the growing need for KRT and the scarce resources available [ 17 ]. Various regions of the country have adapted local variants of this rationing policy, reflecting local resource constraints. These rationing policies’ ability to enable KRT access based on transplantability and potential to reinforce existing socioeconomic inequalities have been controversial. Furthermore, there are scarce data on the ESKD incidence in many SSA countries and the extent to which the KRT need is satisfied. Our study aimed to define the profile of the patients needing KRT, the KRT access rate, and the factors associated with the denial of KRT access in the main public sector hospital of the Free State (FS) Province, RSA.

Methods

Patients and study design

In this retrospective study, we reviewed patients’ records that were referred for KRT and were evaluated by the KRT committee at Universitas Academic Hospital (UAH) from 1 January 2016 to 31 December 2018. Patients evaluated by the KRT committee included those with CKD stages 4 and 5 according to eGFR and those assessed by the referring health care practitioner to need dialysis due to symptoms or biochemical abnormalities attributed to chronic renal failure. UAH is an academic hospital located in Bloemfontein, RSA, and provides tertiary services to the FS province. The kidney unit at UAH receives referrals for KRT from public sector health care facilities throughout the FS province. The FS province is a predominantly rural province with 2.9 million people according to the 2018 estimates, and it covers an area of 129 825 km², with approximately 85 % of the population served by the public health sector [ 18 , 19 ]. The province has 17 dialysis centres: 11 privately owned centres serving those with health insurance and 6 public sector dialysis centres coordinated centrally by the UAH. The FS public health care sector served an estimated 2.36, 2.44, and 2.48 million people without medical insurance in 2016, 2017, and 2018, respectively [ 19 , 20 ]. The FS KRT programme offers peritoneal dialysis (PD) preferentially to patients starting dialysis and haemodialysis to patients unable to undergo PD. At the same time, kidney transplantation is limited to patients with live related donors who can travel to neighbouring provinces for transplantation.

The UAH KRT committee, which comprises hospital managers, nephrologists, attending medical officers, social workers, dieticians, and nursing staff, meet weekly to discuss dialysis-requiring patients referred to the unit. Patients are assessed for KRT eligibility according to the FS KRT guidelines [ 21 ]. Patient considered as not the best candidates for kidney transplantation are classified as category 3 and are not offered KRT. Those with comorbidities that are manageable and potentially reversible are classified as category 2 and are offered KRT when dialysis resources are available. Category 1 includes young patients considered good transplant candidates with minimal comorbidities and are prioritised for KRT (Fig. 1 ). Fig. 1 Free State chronic kidney replacement therapy guidelines. ARV, antiretroviral; BMI, body mass index; HIV, human immunodeficiency virus; LVEF, left ventricular ejection fraction; RVPSP, right ventricular peak systolic pressure

The diaries containing minutes of the KRT committee meetings, the assessment forms, national health laboratory service electronic results, and electronic hospital records (MEDITECH) were reviewed to obtain the names and demographic details, KRT committee discussion outcomes, laboratory test results, and clinical information. The study population comprised adult patients (≥ 18 years old) diagnosed with ESKD and evaluated by the UAH KRT committee between 2016 and 2018. Patients with incomplete data in the assessment tool documents and clinical and socio-demographic records were excluded.

All category 3 patients were denied KRT access according to the FS KRT policy, and both categories 1 and 2 were deemed acceptable for KRT access. Criteria for exclusion from KRT access included concomitant irreversible advanced organ dysfunction. Cardiac disease was assessed using an echocardiogram and defined as an ejection fraction of < 45 % or diastolic dysfunction grade 2 or higher. Significant vascular disease was defined by any vascular calcifications that can be visualised on X-ray. Pulmonary hypertension was defined as a right ventricular peak systolic pressure of > 50 mmHg. Other classifying criteria are depicted in Fig. 1 .

The incidence rates were calculated as the number of new cases divided by the provincial population served by the public health care sector as published by Statistics South Africa for that period. The rates were expressed as per million population (pmp).

Statistical analysis

Continuous variables are summarised using medians and interquartile ranges and compared using the Wilcoxon-Mann-Whitney test. Kruskal-Wallis equality-of-populations rank test was used for comparing more than two continuous variables. Categorical and ordinal variables are summarised using proportions and percentages and compared using Pearson’s chi-square test or Fisher’s exact test, as appropriate. Logit transform was used to calculate confidence intervals of proportions and differences compared using a two-sample test for proportions. Univariate logistic regression was used to estimate the association between exclusion from the KRT programme and various risk factors. Multivariable logistic regression analysis was used to identify independent predictors for exclusion from the KRT programme. Covariates included in the logistic regression model were age, sex, employment status, presence of dependants, poor home circumstances as determined by a social worker, diabetes, hypertension, human immunodeficiency virus (HIV) status, history of smoking, and serum creatinine level. All analyses were performed using Stata Statistical Software, Release 15 (StataCorp, College Station, TX, US), and the level of significance was set at p < 0.05.

Resutls

Patient Characteristics and Demographics

From January 2016 to December 2018, a total of 363 patients were assessed for KRT eligibility at the UAH, 96 patient files were excluded for incomplete information, leaving 267 files for further analysis. Only one third (30.7 %; 82/267) of the patients with ESKD, referred for KRT, were offered dialysis, while the rest (69.3 %; 185/267) were denied KRT access. The median age of the patients in our study was 40 (interquartile range [IQR], 33‒49) years. The median age among the patients that were denied KRT access was 43 (IQR 34‒51) compared to 37 (IQR 30‒43) years among those accepted for KRT ( p < 0.001). There were significantly more male patients (56.2 %; 150/267) than female patients (43.8 %; 117/267) ( p = 0.004) in the study cohort. The proportion of employed patients was 47.6 % (126/267), while 44.9 % (119/267) were unemployed, and 6.8 % (18/267) were pensioners. Those with dependants made up 75.5 % (200/267) of our study population, while 21.9 % had no dependants ( p < 0.001). Only 5.6 % (15/267) of the patients admitted to smoking and alcohol use, which was significantly higher among those excluded from KRT access (7.6 % vs. 1.2 %; p = 0.043). According to the social worker assessment, slightly less than 2 % of the patients were considered to have poor home circumstances and poor social support networks (Table 1 ). Table 1 Background characteristics of end-stage kidney disease patients presenting for kidney replacement therapy KRT Accepted ( n = 82) KRT excluded ( n = 185) p -value Total (%, 95 % CI) p -value Age (years), median (IQR) 37 (30–43) 43 (34–51) < 0.001 40 (33–49) Sex, n (%) Male 46 (56.1) 104 (56.2) 0.986 150 (56.2; 50.0–62.0) 0.004 Female 36 (43.9) 81 (43.8) 117 (43.8; 38.0–49.9) Employment, n (%) Employed/Self employed 40 (49.4) 86 (46.7) 0.691 126 (47.6; 41.6–53.6) Reference Unemployed 40 (49.4) 79 (42.9) 0.331 119 (44.9; 39.0–51.0) 0.542 Pensioner 0 18 (9.8) 0.002 18 (6.8; 4.3–10.5) < 0.001 Unspecified 1 (1.2) 1 (0.5) 0.519 2 (0.8; 0.1–3.0) < 0.001 Marital status, n (%) Married 41 (50.6) 105 (57.1) 0.331 146 (55.1; 49.0–61.0) Reference Single 40 (49.4) 71 (38.6) 0.101 111 (41.9; 36.1–47.9) 0.002 Divorced 0 6 (3.3) 0.182 6 (2.3; 1.0–5.0) < 0.001 Unknown 0 2 (1.1) 1.000 2 (0.8; 0.1–3.0) < 0.001 Dependants, n (%) Yes 56 (69.1) 144 (78.3) 0.112 200 (75.5; 70.0–80.3) Reference No 21 (25.9) 37 (20.1) 0.291 58 (21.9; 17.3–27.3) < 0.001 Unknown 4 (4.9) 3 (1.6) 0.206 7 (2.6; 1.3–5.4) < 0.001 Citizenship, n (%) South African 80 (97.6) 178 (96.2) 0.726 258 (96.6; 93.6–98.2) < 0.001 Non-South African 2 (2.4) 7 (3.8) 9 (3.4; 1.8–6.4) Smoking and/or alcohol, n (%) 1 (1.2) 14 (7.6) 0.043 15 (5.6; 3.4–9.1) Illicit substance abuse, n (%) 0 2 (1.1) 1.000 2 (0.8; 0.2–3.0) Social factors, n (%) Poor home circumstances 1 (1.2) 4 (2.2) 0.548 5 (1.9; 0.8–4.4) 1.000 Poor social network/support 0 5 (2.8) 5 (1.9;0.8–4.4) CI confidence interval; IQR interquartile range; KRT kidney replacement therapy *Illicit substance abuse includes regular use of cannabis and other illegal substances

Clinical characteristics

Among the patients with ESKD included in the study, the most prevalent comorbidities were hypertension (42.3 %; 113/267), HIV (28.5 %; 76/267) and diabetes mellitus (19.1 %; 51/267). Likewise, the three most significant causes of ESKD were hypertension (32.2 %; 86/267), HIV-associated nephropathy (15.7 %; 42/267) and diabetes mellitus (12.4 %; 33/267). Glomerulonephritis was documented as the cause of ESKD in 9 % of the study patients. The median serum creatinine at presentation was 890 (IQR, 670‒1210) µmol/L, with the accepted group having a higher median (1044.5; IQR 807‒1403 µmol/L) than that in the denied group (794; IQR, 587‒1143 µmol/L), p < 0.001 (Table 2 ). Table 2 Clinical characteristics KRT Accepted ( n = 82) KRT excluded ( n = 185) p -value Total (%, 95 % CI) p -value Hypertension, n (%) 36 (43.9) 77 (41.6) 0.728 113 (42.3, 36.5–48.4) Reference Diabetes, n (%) 4 (4.9) 47 (25.4) < 0.001 51 (19.1, 14.8–24.3) < 0.001 SLE, n (%) 6 (7.3) 6 (3.2) 0.138 12 (4.5, 2.6–7.8) < 0.001 Hepatitis B, n (%) 1 (1.2) 4 (2.2) 1.000 5 (1.9, 0.7–4.4) < 0.001 Hepatitis C, n (%) 1 (1.2) 1 (0.5) 0.521 2 (0.8, 0.2–3.0) < 0.001 HIV status, n (%) Negative 61 (74.4) 113 (61.1) Reference 174 (65.2, 59.2–70.7) Positive 21 (25.6) 55 (29.7) 0.491 76 (28.5, 23.4–34.2) 0.001 Unknown 0 17 (9.2) 0.002 17 (6.4, 4.0–10.0) Kidney disease aetiology, n (%) Hypertension 27 (32.9) 59 (31.9) 0.867 86 (32.2, 26.8–38.1) Reference Diabetes mellitus 1 (1.2) 32 (17.3) < 0.001 33 (12.4, 8.9–16.9) < 0.001 Glomerulonephritis 11 (13.4) 13 (7.0) 0.092 24 (9.06.1–13.1) < 0.001 Lupus nephritis 6 (7.3) 6 (3.2) 0.138 12 (4.5,2.6–7.8) < 0.001 HIVAN 14 (17.1) 28 (15.1) 0.688 42 (15.7, 11.8–20.6) < 0.001 FSGS 1 (1.2) 2 (1.1) 1.000 3 (1.1, 0.3–3.4) < 0.001 Autosomal Polycystic kidney disease 6 (7.3) 5 (2.7) 0.098 11 (4.1, 2.3–7.3) < 0.001 Connective tissue diseases/Vasculitis 0 5 (2.7) 0.328 5 (1.9, 0.8–4.4) < 0.001 Obstructive uropathy 1 (1.2) 10 (5.4) 0.181 11 (4.1, 2.3–7.3) < 0.001 Other a 2 (2.4) 7 (3.8) 0.726 9 (3.4, 1.8–6.4) < 0.001 Unknown 13 (15.9) 18 (9.7) 0.150 31 (11.6, 8.3–16.1) < 0.001 Serum creatinine (µmol/L), median (IQR) 1044.5 (807–1403) 794.5 (587–1143) < 0.001 890 (670–1210) CI confidence interval; ESKD end-stage kidney disease; FSGS focal segmental glomerulosclerosis; HIV human immunodeficiency virus; HIVAN HIV associated nephropathy; IQR interquartile range; pmp, per-million-population; KRT kidney replacement therapy; SLE systemic lupus erythematosus a Other causes included oxalosis, multiple myeloma, reflux nephropathy, hypoplastic kidneys and renal agenesis

ESKD incidence and KRT rates

The ESKD incidence rate was 26.7 pmp in 2016, 41.4 pmp in 2017, and 41.6 pmp in 2018, with a cumulative incidence rate of 110.1 pmp over these three years. KRT access rates were 30.2 % (19/63), 27.7 % (28/101), and 34.0 % (35/103) of study patients referred for KRT access in 2016, 2017 and 2018, respectively ( p = 0.622). When all 363 patients referred for KRT at the UAH between 2016 and 2018 were considered, the cumulative ESKD incidence rate was 149.7 (95 % CI, 125.2‒174.1) pmp with an estimated average annual ESKD incidence of 49.9 (95 % CI, 35.8‒64.0) pmp. A steady increasing trend of KRT incidence rate was observed, starting at 8.05 (95 % CI, 4.98‒11.1) in 2016, 11.5 (95 % CI, 7.83‒15.1) in 2017 and 14.1 (95 % CI, 10.3‒18.0) in 2018 (Table 3 ). The KRT incidence increased by 75 % between 2016 and 2018 ( p = 0.045). Table 3 Access to kidney replacement therapy 2016 ( n = 63) 2017 ( n = 101) 2018 ( n = 103) p -value Total (%, 95 % CI) ( n = 267) p -value ESKD incidence rate, pmp (95 % CI) 26.7 (20.0–33.4) 41.4 (33.2–49.6) 41.6 (33.4–49.8) 0.006 a 0.005 b 110.1 (96.6–123.6) Accepted for dialysis, n (%) 19/63 (30.2 %; 20.0–42.8 %) 28/101 (27.7 %; 19.8–37.3 %)) 35/103 (34.0 %; 25.4–43.7 %)) 0.622 82/267 (30.7 %; 25.4–36.5 %)) KRT incidence, pmp (95 % CI) 8.05 (4.98–11.1) 11.5 (7.83–15.1) 14.1 (10.3–18.0) 0.230 a 0.045 b 33.8 (27.7–39.9) Age (years), median (IQR) 36 (32–48) 43 (35–50) 40 (31–48) 0.131 40 (33–49) Hypertension, n (%) 27/63 (42.8) 50/101 (49.5) 36/103 (35.0) 0.109 113/267 (42.3, 36.5–48.4) Diabetes, n (%) 13/63 (20.6) 21/101 (20.8) 17/103 (16.5) 0.693 51/267 (19.1, 14.8–24.3) HIV, n (%) 13/63 (20.6) 30/101 (29.7) 33/103 (32.0) 0.270 76/267 (28.5, 23.4–34.2) Serum creatinine (µmol/L), median (IQR) 807 (529–1050) 995.5 (708–1050) 883 (650–1184) 0.033 890 (670–1210) Reason for exclusion from KRT programme Concomitant organ dysfunction, n (%) 25/44 (56.8) 45/73 (61.6) 39/68 (57.4) 0.830 109/185 (58.9; 51.6–65.8) Reference Cardiac disease, n (%) 9/44 (20.4) 16/73 (21.9) 22/68 (32.4) 0.250 47/185 (25.4;19.6–32.2) Reference Vascular disease, n (%) 7/44 (20.4) 8/73 (23.3) 12/68 (16.2) 0.511 27/185 (14.6; 10.2–20.5) 0.009 Pulmonary hypertension, n (%) 9/44 (20.4) 17/73 (23.3) 11/68 (16.2) 0.571 37/185 (20.0; 14.8–26.4) 0.215 Lung disease, n (%) 1/44 (2.3) 5/73 (6.8) 2/68 (2.9) 0.467 8/185 (4.3; 2.2–8.4) < 0.001 Noncompliance, n (%) 4/44 (9.1) 2/73 (2.7) 4/68 (5.9) 0.286 10/185 (5.4; 2.9–9.8) < 0.001 Psychosocial, n (%) 2/44 (4.6) 2/73 (2.7) 4/68 (5.9) 0.660 8/185 (4.3; 2.2–8.4) < 0.001 Substance abuse, n (%) 2/44 (4.6) 4/73 (5.5) 1/68 (1.5) 0.483 7/185 (3.8; 1.8–7.8) < 0.001 Diabetes and age ≥ 50 years, n (%) 3/44 (6.8) 9/73 (12.3) 2/68 (2.9) 0.124 14/185 (7.6; 4.5–12.4) < 0.001 Uncontrolled HIV c , n (%) 3/44 (6.8) 9/73 (12.3) 9/68 (13.2) 0.610 21/185 (11.4; 7.5–16.8) < 0.001 Age ≥ 60 years, n (%) 4/44 (9.1) 7/73 (9.6) 6/68 (8.8) 1.000 17/185 (9.2; 5.8–14.3) < 0.001 BMI ≥ 35 kg/m 2 , n (%) 3/44 (6.8) 3/73 (4.1) 7/68 (10.3) 0.357 13/185 (7.0; 4.1–11.8) < 0.001 Other, n (%) 2/44 (4.6) 3/73 (4.1) 4/68 (5.9) 0.910 9/185 (4.9; 2.5–9.1) < 0.001 BMI body mass index; CI confidence interval; ESKD end-stage kidney disease; HIV human immunodeficiency virus; IQR interquartile range; pmp per-million-population; KRT kidney replacement therapy a Comparison of 2016 and 2017 incidence rates b Comparison of 2016 and 2018 incidence rates c Uncontrolled HIV refers to unsuppressed viral load

Reasons for non-admission to the KRT program

The main reasons for being denied KRT access in the FS public sector are listed in Table 3 . Medical factors accounted for the largest proportion of reasons to be denied KRT, including communicable and non-communicable diseases. Concomitant advanced organ dysfunction was the most typical medical cause of denial (58.9 %; 109/185), including cardiac disease (25.4 %; 47/185), vascular disease (14.6 %; 27/185), chronic lung disease (4.3 %; 8/185), among others. Uncontrolled HIV infection (11.4 %; 21/185) and age above 60 years (9.2 %; 17/185), respectively, were other prominent exclusion factors. Behavioural and psychosocial factors, such as habitual non-compliance to medical treatment, substance abuse, and mental illness, were cited as contributing to 13.5 % of the exclusions combined. Diabetes mellitus was a significantly higher comorbidity among those denied KRT compared to that in patients who were offered KRT (25.4 vs. 4.9 %; p < 0.001).

Age (Odds Ratio [OR], 1.04; 95 % CI, 1.00‒1.07; p = 0.024) and diabetes (OR, 5.04; CI, 1.69‒15.03; p = 0.004) were independent predictors for non-acceptance to the KRT programme. However, hypertension (OR, 1.80; 1.06‒3.04; p = 0.029) independently predicted concomitant advanced organ dysfunction resulting in exclusion from the KRT programme (Table 4 ). Table 4 Univariate and multivariable logistic regression for exclusion from KRT programme Univariate logistic regression Multivariable logistic regression Odds Ratio (95 % CI) p -value Odds Ratio (95 % CI) p -value All-cause exclusion from KRT programme Age 1.04 (1.02–1.07) < 0.001 1.04 (1.00–1.07) 0.024 Sex 1.00 (0.59–1.68) 0.986 0.86 (0.47–1.56) 0.618 Employment status 1.31 (0.86–1.99) 0.202 1.57 (0.96–2.59) 0.075 Has dependents 1.61 (0.89–2.89) 0.113 0.94 (0.41–2.17) 0.888 Poor home circumstances 3.35 (0.59–19.02) 0.172 3.07 (0.50–18.69) 0.224 Diabetes 6.64 (2.31–19.13) < 0.001 5.04 (1.69–15.03) 0.004 Hypertension 0.98 (0.56–1.70) 0.937 0.97 (0.52–1.78) 0.910 HIV 1.23 (0.68–2.21) 0.492 1.12 (0.57–2.20) 0.738 Smoking and alcohol 6.63 (0.86–51.3) 0.070 4.56 (0.56–36.97) 0.155 Serum creatinine 1.00 (1.00–1.00) 0.001 1.00 (1.00–1.00) 0.007 Exclusion due to advanced organ dysfunction. Age 1.00 (0.98–1.02) 0.673 1.00 (0.98–1.02) 0.811 Sex 0.79 (0.48–1.29) 0.345 0.83 (0.49–1.40) 0.486 Hypertension 1.76 (1.07–2.88) 0.026 1.80 (1.06–3.04) 0.029 Diabetes 1.67 (0.90–3.08) 0.103 1.65 (0.85–3.20) 0.136 HIV 0.58 (0.33–1.01) 0.054 0.58 (0.32–1.05) 0.070 Smoking and alcohol 2.28 (0.79–6.60) 0.129 1.97 (0.65–5.95) 0.228 Serum creatinine 1.00 (1.00–1.00) 0.005 1.00 (1.00–1.00) 0.005 CI confidence interval; HIV human immunodeficiency virus; KRT kidney replacement therapy

Discussion

This study describes the epidemiology and KRT access of patients with ESKD referred for dialysis in the public health sector of the FS province, a predominantly rural province in South Africa. An average annual ESKD incidence of 49.9 pmp was estimated, with 30 % of those presenting with ESKD able to access KRT. Hypertension, diabetes, and HIV were the commonest comorbidities, while advanced organ dysfunction and uncontrolled HIV were the leading reasons for denying KRT access. Age and diabetes mellitus were independent predictors for denial of KRT access.

The crude annual incidence rates of ESKD increased from 26.7 to 41.6 pmp between 2016 and 2018. However, the crude average annual ESKD incidence rate was 49.9 pmp when the entire cohort of 363 patients was considered, including the 96 patient records excluded from analysis due to incomplete data. Still, this is likely underestimating the true ESKD incidence rate in the FS since some patients were not referred to the UAH, particularly if they had an obvious exclusion factor. The ESKD incidence reported in the literature is highly variable, ranging from 89 pmp in some western countries such as Finland to 363 pmp in the United States of America [ 22 ]. In Latin America, a high incidence rate of 458 pmp has been reported in Mexico [ 23 ]. There is a dearth of reported ESKD incidence rates in Africa due to the limited access to KRT and poor data collection mechanisms [ 13 ]. With relatively non-restrictive KRT access and a high prevalence of associated non-communicable diseases, Libya reported an ESKD incidence of 282 pmp [ 24 ]. These higher incidence rates than those estimated in this study suggest a significant underestimation of ESKD incidence in RSA, as a substantial number of patients with ESKD are palliated at the primary health care level without referral to a specialised kidney unit.

A typical patient with ESKD referred for KRT was young (mean age of 40 years old), married with dependents, and of male preponderance. This young profile was primarily attributed to the KRT policy excluding those with age above 60 years and age above 50 years in people with diabetes, thus influencing the likelihood of referral from primary health care levels. Other contributors included the favourable consideration of society’s economically active members. Furthermore, age was a significant predictor of KRT access, with those offered KRT access being significantly younger than those excluded (37 vs. 43 mean years old, p < 0.001). Moreover, the increased likelihood of having comorbidities and advanced organ damage with progressing age was another reason for the age-associated bias. Typically, advancing age is associated with an increased incidence of ESKD [ 22 , 25 ]. Thus, the age criterion likely contributed to a significant undermining of the ESKD incidence, since elderly patients are being censored at the primary health care level.

Regarding behavioural factors, the use of alcohol and smoking was significantly higher in the excluded group. According to the program guidelines, no one was excluded from the KRT program solely based on smoking or alcohol use. Nonetheless, these behavioural factors were present disproportionally in some patients excluded due to advanced organ dysfunction and substance abuse, including alcohol abuse. Social factors such as poor home circumstances and social support networks were not significantly higher in the excluded group. However, Cape Town reports reviewing the KRT programme in the Western Cape (WC) province, RSA, showed a significant influence of psychosocial factors in denying KRT access [ 11 , 26 ]. This incongruity highlights differing tolerances and thresholds in evaluating such highly subjective social variables among the country’s different regions. The FS is predominantly a rural province with a significant indigent population and a gross domestic product lower than half that of WC. Nearly half of the adults in FS live below $2.5/day compared to a third in the WC [ 27 ]. Thus, it is conceivable that the FS employed a much more accommodative approach to determine what constitutes poor home circumstances and social support compared to that in the WC. A more liberal outlook would accommodate the expected higher proportion of FS residents living under rural and poor socioeconomic conditions.

At the time of the KRT eligibility assessment, the mean creatinine level was documented as 890 µmol/L, with the accepted group having a significantly higher creatinine level than that in the excluded group (1044 vs. 794 µmol/L, p < 001). The lower creatinine levels among the excluded group were likely an artefact of non-referral of those with severe kidney failure and an obvious exclusion factor. These would have been palliated at the primary health care level if there were no suggestions of acute kidney failure. Those with suspicion of acute kidney disease were offered dialysis for up to 6 weeks. High levels of creatinine at referral suggest a prevailing problem of late presentation of patients with advanced CKD to the nephrology unit. Failure of early identification of CKD and timely referral for tertiary kidney care highlights a primary health care deficiency in the FS public health care system and RSA. A late referral is associated with adverse morbidity and mortality outcomes leading to unfavourable socioeconomic consequences, which can be counterproductive in a resource-limited environment [ 28 , 29 ]. A shortage of nephrologists, most of whom are concentrated in urban centres, and poor awareness of CKD among primary health care providers are likely underlying factors that contribute to the late referral of ESKD [ 15 , 30 ]. The diverse and immense patient burdens faced by the few primary health care providers in rural and peri-urban health care centres constitute another possible contributing factor to inadequate surveillance and timely identification of CKD.

Prevention and early CKD identification are particularly important among high-risk groups such as those with hypertension, diabetes, and HIV [ 31 ]. Screening for proteinuria and reductions in GFR are cost-effective measures that allow for early diagnosis and intervention to reduce the risk of ESKD and the associated cardiovascular events and mortality [ 32 ]. Although population-wide urinalysis screening of working adults and school-aged children aided in lowering the incidence of glomerulonephritis and ESKD in Japan [ 33 ], the cost-effectiveness of mass screening strategies in resource-limited low-to-middle-income countries is controversial [ 31 , 34 ]. Both primary and secondary prevention strategies depend on the available local resources [ 35 ]. Adequately trained primary healthcare personnel who can recognise and manage both communicable and non-communicable diseases is critical for a sustainable prevention strategy. The RSA public health sector is fortunate to have free and general availability of major drugs needed to manage hypertension, diabetes, HIV, and other common disorders. This essential drug availability includes drugs vital to early CKD care, such as inhibitors of the renin-angiotensin-aldosterone system. However, shortages of appropriately capacitated human capital are crucial barriers that impede the success of prevention strategies in the FS and RSA as a whole [ 15 ]. Investing in the capacitation of the public health care system and promoting clinical practice standards that encourage regular screening and initiation of early interventions against kidney injury, communicable diseases, and non-communicable diseases, among others, are vital in mitigating the risk posed by ESKD.

Both communicable and non-communicable diseases were associated with the reported cause of kidney disease in our cohort. Hypertension and HIV were the most prevalent comorbidities and commonest attributed causes of ESKD, with no significant differences between the accepted and the non-accepted groups. Likewise, hypertension has been reported as the commonest cause of ESKD among those receiving KRT in the RSA [ 36 ]. Furthermore, hypertension has been noted to be the foremost cause of kidney disease in SSA [ 13 ]. However, physician-reported ESKD causes attributed to hypertension are notoriously unreliable, as hypertension is both a cause and a consequence of ESKD, especially when the presentation is late and without the benefit of a kidney biopsy. By contrast, diabetes has been reported to be the most common cause of ESKD in both United States (USRDS) and European (ERA-EDTA) kidney registries [ 25 , 37 ]. In our cohort, diabetes mellitus was the third-commonest cause of ESKD and was significantly more common among the excluded group than among the KRT-offered group. However, this result was likely influenced by the non-referral sorting of patients with diabetes who are older than 50 years, as it is a recognised exclusion parameter. This bias likely resulted in an underappreciation of diabetes-associated kidney disease in our cohort. The prevalence of diabetes and its associated complications such as kidney diseases is expected to increase, with SSA projected to have the largest increase [ 38 , 39 ]. As diabetes and the related ESKD are linked to significant cardiovascular morbidity and mortality, considerable associated socioeconomic and public health burdens can be imposed on societies who can least afford the additional burden [ 40 , 41 ].

Given the shortage of dialysis machines and health care personnel and the high cost associated with KRT, the FS KRT rationing programme attempts to accommodate those who can potentially benefit the most from KRT. During the study, the KRT incidence increased by 75 %, from 8.05 pmp in 2016 to 14.2 pmp in 2018. However, the acceptance rate did not change significantly during the same period, with two-thirds of the referred patients with ESKD not being offered dialysis. Even with the reported underestimated burden of ESKD, it is evident that the FS KRT program cannot support the indigent population’s KRT needs. A systemic review by Liyanage et al. estimated Africa to have the lowest KRT access rates in the world, ranging from 9 to 16 % [ 14 ]. Considering those who were not referred for KRT in UAH, it is plausible that the FS’s true KRT access rate is close to Africa’s projected range. This dire situation highlights the need for RSA to invest in cost-effective ways to increase KRT access progressively; otherwise, efforts directed at early diagnosis would be futile in those who eventually progress to ESKD. Further, the early awareness of CKD and the realisation of limited access to KRT when needed produce an appreciable emotional toll on the health caregiver, patient, and family.

On logistic regression analysis, age and diabetes mellitus were shown to be independent predictors of exclusion from the KRT program. These two factors are associated with significant morbidity and mortality in patients on KRT, and thus carry an extra-economic burden on the health care system [ 42 ]. Thus, exclusion based on these factors avoids the added risk levied by them. This utilitarian outlook is one approach to addressing the ethical dilemma faced by many SSA countries of increasing health care demands in an environment with scarce prevailing resources. However, denying KRT access to those who may have contributed the most to society over the years while favouring the young who can still contribute economically is ethically challenging. Furthermore, as the FS and RSA KRT programme is based on transplantability with age no longer a widely recognised contraindication to transplantation globally, this exclusion factor is a contradiction [ 43 ]. Reports from transplant programmes with universal access show that older adults benefit significantly in terms of survival and quality of life from transplantation, with the average age of transplantation at above 60 years in many western countries [ 25 , 44 , 45 ]. Although some reports have noted increased mortality and hospitalisation rates associated with age, the reported graft and patient survival rates are more than 90 and 77 % at 1 and 5 years, respectively [ 46 – 48 ]. The age contradiction has persisted due to RSA’s resource constraints, as in many SSA countries.

Concomitant advanced and irreversible organ dysfunction, including cardiovascular and cardiopulmonary diseases, was the commonest reason for exclusion from the KRT programme. Cardiac disease and pulmonary hypertension were the commonest advanced organ dysfunctions cited as reasons for KRT access denial. Hypertension was identified as a significant predictor of concomitant advanced organ dysfunction resulting in exclusion. This association is likely a consequence of cardiovascular complications associated with hypertension. Likewise, cardiovascular complications in ESKD are associated with increased morbidity and mortality and can potentially impose additional socioeconomic risk on the health care system. Uncontrolled HIV, defined by unsuppressed viral load, was the second reason cited for denial of KRT. This exclusion is based on the high morbidity and mortality associated with uncontrolled HIV among ESKD patients on dialysis [ 49 – 51 ]. In 2016, the RSA government introduced a universal test-and-treat policy, making available antiretroviral treatment to all who test positive for HIV irrespective of CD4 count. This strategy is hoped to progressively mitigate against the complications, morbidity and mortality associated with HIV infection [ 52 ]. Likewise, these universal provisions should alleviate the presentation of patients with ESKD and uncontrolled HIV, as this strategy’s benefits are gradually realised.

Our study’s main limitation was that it was a single-centre retrospective study with several incomplete files being excluded in the analysis, thereby limiting the generalisability of the findings. Some of the incomplete files were of patients who died shortly after admission to the UAH before a decision on the eligibility for KRT could be made. The non-referral of some patients with ESKD to the UAH with obvious exclusion factors likely contributed to the underestimation of ESKD incidences and KRT access rates in the FS. However, this study’s strength is that the UAH is the only centre that processes KRT access for the entire FS province, allowing for KRT incidence rates of the province’s public sector to be readily attainable.

Conclusions

The FS public health care sector serves a mostly young and economically able but indigent portion of the society. Both non-communicable and communicable diseases, including hypertension, diabetes, and HIV, contributed to the causes of ESKD. Although two-thirds of the patients are still excluded from accessing KRT, the FS provincial KRT incidence grew by 75 % between 2016 and 2018. Age, diabetes mellitus, and advanced organ dysfunction were significant factors for KRT access denial. Hypertension was the major independent risk factor for developing concomitant advanced organ dysfunction resulting in exclusion from dialysis access. All major ESKD aetiologies in our study were potentially preventable and manageable. Strategies aimed at early identification and prevention at the primary health care level may help alleviate some of the burdens imposed by ESKD on the RSA public health system. However, more resources are needed to meet the increasing demands for KRT. Likewise, more country-level studies are required to define further the epidemiology of CKD and contributors to ESKD in RSA.

Supplementary Information

Additional file 1:

Abbreviations

End-stage kidney disease

Confidence interval

Chronic kidney disease

European Renal Association – European Dialysis and Transplant Association

Free State province

Glomerular filtration rate

Human immunodeficiency virus

Interquartile range

Odds ratio

Peritoneal dialysis

Per million population

Kidney replacement therapy

Republic of South Africa

Sub-Saharan Africa

Universitas Academic Hospital

United States Renal Data System

Western Cape province

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Thabang T Molaoa, Feziwe B Bisiwe and Kwazi CZ Ndlovu contributed equally to this work.

Acknowledgements

The authors wish to thank Dr Claire Amour Barrett and all the Universitas Academic Hospital Renal unit staff who helped with the data collection.

Authors’ contributions

Research idea and study design: TM, FB, KN; data acquisition: TM; data analysis and interpretation: KN, FB, TM; statistical analysis: KN, FB, TM; supervision and mentoring: FB, KN. Each author contributed important intellectual content during manuscript drafting or revision and accepts accountability for the overall work by ensuring that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved. All authors have read and approved the final manuscript.

Funding

No external funding was received.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the University of the Free State Health Sciences Research Ethics Committee (UFSHSD2019/1083), and research was conducted in accordance with the principles of the Declaration of Helsinki. Informed consent from the participants was not required owing to the retrospective nature of the study.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References

1. Levey AS Eckardt KU Tsukamoto Y Levin A Coresh J Rossert J Definition and classification of chronic kidney disease: a position statement from kidney disease: improving global outcomes (KDIGO) Kidney Int 2005 67 2089 100 10.1111/j.1523-1755.2005.00365.x 15882252

2. Naicker S Burden of end-stage renal disease in sub-Saharan Africa Clin Nephrol 2010 74 Suppl 1 13 6 10.5414/cnp74s013 20979956

3. Murphy D McCulloch CE Lin F Banerjee T Bragg-Gresham JL Eberhardt MS Trends in prevalence of chronic kidney disease in the United States Ann Intern Med. 2016 165 473 81 10.7326/M16-0273 27479614 PMC5552458

4. Zhang QL Rothenbacher D Prevalence of chronic kidney disease in population-based studies: systematic review BMC Public Health. 2008 8 117 10.1186/1471-2458-8-117 18405348 PMC2377260

5. Hill NR Fatoba ST Oke JL Hirst JA O’Callaghan CA Lasserson DS Global prevalence of chronic kidney disease - a systematic review and meta-analysis PLoS One 2016 11 e0158765 10.1371/journal.pone.0158765 27383068 PMC4934905

6. Hodel NC Hamad A Praehauser C Mwangoka G Kasella IM Reither K The epidemiology of chronic kidney disease and the association with non-communicable and communicable disorders in a population of sub-Saharan Africa PLoS One 2018 13 e0205326 10.1371/journal.pone.0205326 30379902 PMC6209178

7. Arogundade FA Barsoum RS CKD prevention in Sub-Saharan Africa: a call for governmental, nongovernmental, and community support Am J Kidney Dis 2008 51 515 523 10.1053/j.ajkd.2007.12.006 18295068

8. Stanifer JW Jing B Tolan S Helmke N Mukerjee R Naicker S The epidemiology of chronic kidney disease in sub-Saharan Africa: a systematic review and meta-analysis Lancet Glob Health 2014 2 e174-81 10.1016/S2214-109X(14)70002-6 25102850

9. Mayosi BM Flisher AJ Lalloo UG Sitas F Tollman SM Bradshaw D The burden of non-communicable diseases in South Africa Lancet 2009 374 934 47 10.1016/S0140-6736(09)61087-4 19709736

10. Etheredge H Fabian J Challenges in expanding access to dialysis in South Africa—expensive modalities, cost constraints and human rights Healthcare 2017 5 38 10.3390/healthcare5030038 PMC5618166 28758959

11. Kilonzo KG Jones ESW Okpechi IG Wearne N Barday Z Swanepoel CR Disparities in dialysis allocation: An audit from the new South Africa PLoS One 2017 12 e0176041 10.1371/journal.pone.0176041 28419150 PMC5395209

12. Wang V Vilme H Maciejewski ML Boulware LE The economic burden of chronic kidney disease and end-stage renal disease Semin Nephrol 2016 36 319 30 10.1016/j.semnephrol.2016.05.008 27475662

13. Naicker S End-stage renal disease in Sub-Saharan Africa Kidney Int Suppl 2013 3 161 63 10.1038/kisup.2013.4 12864889

14. Liyanage T Ninomiya T Jha V Neal B Patrice HM Okpechi I Worldwide access to treatment for end-stage kidney disease: a systematic review Lancet 2015 385 1975 82 10.1016/S0140-6736(14)61601-9 25777665

15. Hassen M Archer E Pellizzon A Chikte UME Davids MR Human resources for nephrology in South Africa: A mixed-methods study PLoS One 2020 15 e0228890 10.1371/journal.pone.0228890 32053647 PMC7018074

16. Mayosi BM Benatar SR Health and health care in South Africa–20 years after Mandela N Engl J Med 2014 371 1344 53 10.1056/NEJMsr1405012 25265493

17. Naicker S End-stage renal disease in sub-Saharan and South Africa Kidney Int Suppl 2003 63 119 22 10.1046/j.1523-1755.63.s83.25.x 12864889

18. Statistics SA: Census 2011 Census in brief. Statistics South Africa. 2012. http://www.statssa.gov.za/census/census_2011/census_products/Census_2011_Census_in_brief.pdf . Accessed 9 Feb 2021.

19. Statistics SA: 2016–2018 Mid-year population estimates. Statistics South Africa. 2019. http://www.statssa.gov.za/ . Accessed 5 Feb 2021.

20. Statistics SA: 2016–2018 General Household Survey. Statistics South Africa. 2019. http://www.statssa.gov.za/ . Accessed 5 Feb 2021.

21. Moosa MR. Priority setting approach in the selection of patients in the public sector with end-stage kidney failure for renal replacement treatment in the Western Cape Province. 2013. http://s3.documentcloud.org/documents/19489/moosa-priority-setting-policy-final-feb-24-2010-final.pdf . Accessed on 2nd March 2021

22. De Meyer V, Abramowicz D, De Meester J, Collart F, Bosmans JL, Cools W, et al. Variability in the incidence of kidney replacement therapy over time in Western industrialised countries: A retrospective registry analysis. PLoS One. 2020;15: e0235004. 10.1371/journal.pone.0235004 PMC7316284 32584849

23. Gonzalez-Bedat M Rosa-Diez G Pecoits-Filho R Ferreiro A Garcia-Garcia G Cusumano A Burden of disease: prevalence and incidence of ESRD in Latin America Clin Nephrol Suppl 2015 1 3 6 10.5414/cnp83s003 25725232

24. Alashek WA McIntyre CW Taal MW Epidemiology and aetiology of dialysis-treated end-stage kidney disease in Libya BMC Nephrol 2012 13 33 10.1186/1471-2369-13-33 22682181 PMC3407761

25. Kramer A, Boenink R, Noordzij M, Bosdriesz JR, Stel VS, Beltran P, et al. The ERA-EDTA Registry Annual Report 2017: a summary. Clin Kidney J. 2020;13:693–709. 10.1093/ckj/sfaa048 PMC7467580 32897277

26. Moosa MR Maree JD Chirehwa MT Benatar SR Use of the ‘accountability for reasonableness’ approach to improve fairness in accessing dialysis in a middle-income country PLoS One. 2016 11 e0164201 10.1371/journal.pone.0164201 27701466 PMC5049822

27. Statistics SA. Findings from the Living Conditions Surveys, 2008/2009–2014/2015. Statistics South Africa. 2017. http://www.statssa.gov.za/ . Accessed 5 Feb 2021.

28. Arora P Obrador GT Ruthazer R Kausz AT Meyer KB Jenuleson CS Prevalence, predictors, and consequences of late nephrology referral at a tertiary care center J Am Soc Nephrol 1999 10 1281 6 10.1681/ASN.V1061281 10361866

29. Navaneethan SD Aloudat S Singh S A systematic review of patient and health system characteristics associated with late referral in chronic kidney disease BMC Nephrol 2008 9 3 10.1186/1471-2369-9-3 18298850 PMC2291456

30. Lunney M Alrukhaimi M Ashuntantang GE Bello AK Bellorin-Font E Benghanem Gharbi M Guidelines, policies, and barriers to kidney care: findings from a global survey Kidney Int Suppl 2018 8 30 40 10.1016/j.kisu.2017.10.007 PMC6336223 30675437

31. Vanholder R Annemans L Brown E Gansevoort R Gout-Zwart JJ Lameire N Reducing the costs of chronic kidney disease while delivering quality health care: a call to action Nat Rev Nephrol 2017 13 393 409 10.1038/nrneph.2017.63 28555652

32. James MT Hemmelgarn BR Tonelli M Early recognition and prevention of chronic kidney disease Lancet 2010 375 1296 309 10.1016/S0140-6736(09)62004-3 20382326

33. Imai E Yamagata K Iseki K Iso H Horio M Mkino H Kidney disease screening program in Japan: history, outcome, and perspectives Clin J Am Soc Nephrol 2007 2 1360 6 10.2215/CJN.00980207 17942780

34. Codreanu I Perico N Sharma SK Schieppati A Remuzzi G Prevention programmes of progressive renal disease in developing nations Nephrology 2006 11 321 8 10.1111/j.1440-1797.2006.00587.x 16889572

35. Luyckx VA Cherney DZI Bello AK Preventing CKD in developed countries Kidney Int Rep 2020 5 263 77 10.1016/j.ekir.2019.12.003 32154448 PMC7056854

36. Davids MR, Jardine T, Marais N, Jacobs JC, Sebastian S. South African renal registry annual report 2018. Afr J Nephrol. 2020;23:185 – 96.

37. Saran R, Robinson B, Abbott KC, Bragg-Gresham J, Chen X, Gipson D, et al. US Renal Data System 2019 Annual data report: epidemiology of kidney disease in the United States. Am J Kidney Dis. 2020;Suppl 1:6–7. 10.1053/j.ajkd.2019.09.003 31704083

38. International Diabetes Federation. IDF diabetes atlas 8th edition. 2017. https://www.diabetesatlas.org/upload/resources/previous/files/8/IDF_DA_8e-EN-final.pdf?fbclid=IwAR1dhyFGa3ZlDQngZlEOoDVenpcMPl49ox 8Us8nWL9_rSv0-3PIWcZio8bY. Accessed on 2nd March 2021

39. Mbanya JC Motala AA Sobngwi E Assah FK Enoru ST Diabetes in sub-Saharan Africa Lancet 2010 375 2254 66 10.1016/S0140-6736(10)60550-8 20609971

40. Mapa-Tassou C Katte JC Mba Maadjhou C Mbanya JC Economic impact of diabetes in Africa Curr Diab Rep 2019 19 5 10.1007/s11892-019-1124-7 30680578

41. Anders HJ Huber TB Isermann B Schiffer M CKD in diabetes: diabetic kidney disease versus nondiabetic kidney disease Nat Rev Nephrol 2018 14 361 77 10.1038/s41581-018-0001-y 29654297

42. Ma L Zhao S Risk factors for mortality in patients undergoing hemodialysis: A systematic review and meta-analysis Int J Cardiol 2017 238 151 8 10.1016/j.ijcard.2017.02.095 28341375

43. Chadban SJ, Ahn C, Axelrod DA, Foster BJ, Kasiske BL, Kher V, et al. KDIGO clinical practice guideline on the evaluation and management of candidates for kidney transplantation. Transplantation. 2020;Suppl 1:11–103. 10.1097/TP.0000000000003136 32301874

44. Rao PS Merion RM Ashby VB Port FK Wolfe RA Kayler LK Renal transplantation in elderly patients older than 70 years of age: results from the scientific registry of transplant recipients Transplantation 2007 83 1069 74 10.1097/01.tp.0000259621.56861.31 17452897

45. Griva K Davenport A Harrison M Newman SP The impact of treatment transitions between dialysis and transplantation on illness cognitions and quality of life–A prospective study Br J Health Psychol 2012 17 812 27 10.1111/j.2044-8287.2012.02076.x 22536819

46. Dempster NJ Ceresa CDL Aitken E Kingsmore D Outcomes following renal transplantation in older people: a retrospective cohort study BMC Geriatr 2013 13 79 10.1186/1471-2318-13-79 23883109 PMC3729818

47. Morales JM Marcén R del Castillo D Andres A Gonzalez-Molina M Oppenheimer F Risk factors for graft loss and mortality after renal transplantation according to recipient age: a prospective multicentre study Nephrol Dial Transplant 2012 Suppl 4 iv39-46 10.1093/ndt/gfs544 PMC3526982 23258810

48. Knoll GA Kidney transplantation in the older adult Am J Kidney Dis. 2013 61 790 7 10.1053/j.ajkd.2012.08.049 23261121

49. Ndlovu KC Sibanda W Assounga A Peritonitis outcomes in patients with HIV and end-stage renal failure on peritoneal dialysis: a prospective cohort study BMC Nephrol 2017 18 48 10.1186/s12882-017-0466-0 28158991 PMC5291961

50. Ndlovu KCZ Assounga A Continuous Ambulatory Peritoneal Dialysis in Patients with HIV and End-Stage Renal Failure Perit Dial Int 2017 37 321 330 10.3747/pdi.2016.00165 27935535

51. Rodriguez RA Mendelson M O’Hare AM Hsu LC Schoenfeld P Determinants of survival among HIV-infected chronic dialysis patients J Am Soc Nephrol 2003 14 1307 13 10.1097/01.ASN.0000062963.56513.28 12707399

52. Havlir D Lockman S Ayles H Larmarange J Chamie G Gaolathe T What do the Universal Test and Treat trials tell us about the path to HIV epidemic control? J Int AIDS Soc 2020 23 e25455 10.1002/jia2.25455 32091179 PMC7038879