Skip to main content

How to estimate glomerular filtration rate in sub-Saharan Africa: design and methods of the African Research into Kidney Diseases (ARK) study

Abstract

Background

Chronic kidney disease (CKD) is a substantial cause of morbidity and mortality worldwide with disproportionate effects in sub-Saharan Africa (SSA). The optimal methods to estimate glomerular filtration rate (GFR) and therefore to determine the presence of CKD in SSA are uncertain. We plan to measure iohexol excretion to accurately determine GFR in Malawi, South Africa and Uganda. We will then assess the performance of existing equations to estimate GFR and determine whether a modified equation can better improve estimation of GFR in sub-Saharan Africa.

Methods

The African Research on Kidney Disease (ARK) study is a three-country study embedded within existing cohorts. We seek to enrol 3000 adults > 18 years based on baseline serum creatinine. Study procedures include questionnaires on socio-demographics and established risk factors for kidney disease along with anthropometry, body composition, blood pressure, blood chemistry and urine microscopy and albuminuria. We will measure GFR (mGFR) by plasma clearance of iohexol at 120, 180 and 240 min. We will compare eGFR determined by established equations with mGFR using Bland-Altman plots. We will use regression methods to estimate GFR and compare the newly derived model with existing equations.

Discussion

Through the ARK study, we aim to establish the optimal approach to estimate GFR in SSA. The study has the advantage of drawing participants from three countries, which will increase the applicability of the findings across the region. It is also embedded within established cohorts that have longitudinal information and serial measures that can be used to characterize kidney disease over a period of time. This will help to overcome the limitations of previous research, including small numbers, selected population sub-groups, and lack of data on proteinuria.

The ARK collaboration provides an opportunity for close working partnerships across different centres, using standardized protocols and measurements, and shared bio-repositories. We plan to build on the collaboration for this study for future work on kidney disease in sub-Saharan Africa, and welcome additional partners from across the continent.

Peer Review reports

Background

Chronic kidney disease (CKD) is a substantial cause of morbidity and mortality worldwide with an estimated prevalence of 8–16% [1]. A recent systematic review estimated the prevalence of CKD in sub-Saharan Africa to be around 14% [2]. However, the optimal equation to accurately estimate glomerular filtration rate (GFR) among sub-Saharan African populations is uncertain. This uncertainty is a major barrier to identifying individuals with CKD and to estimation of disease burden.

Worldwide, several equations have been used to estimate GFR from serum creatinine [3,4,5,6,7]. Unfortunately, most of these have been derived from high-income countries with minimal validation in sub-Saharan African populations. Their accuracy at an individual level is limited, particularly in relation to the adjustments for African-American ethnicity, which might be related to variations in muscle mass. Serum creatinine levels at a given level of renal function vary substantially with ethnicity, age, sex, physical activity and nutritional status [8,9,10,11,12,13,14]. Several sources of inaccuracy in estimating GFR have been described, including biological variability in serum creatinine, laboratory induced errors in creatinine measurement and choice of estimating equation [15, 16]. The imprecision of creatinine measurement is more marked at values near the normal range where it is most critical to determine earlier stages of CKD [17,18,19]. Several methods for direct measurement of GFR are available through measurement of clearance of inulin, iothalamate, iohexol, and radio-active agents such as technetium-99 m diethylenetriamine penta-acetic acid (DTPA) and chromium-51 ethylene diamine tetra-acetic acid (EDTA), and 24-h urinary creatinine collection for estimation of creatinine clearance [20]. However, all of these have their challenges. DTPA, EDTA and inulin are expensive and not readily available in most African countries, while 24-h urinary collection is often inaccurate due to difficulty in ensuring a complete sample, coupled with the additional limitation of tubular creatinine secretion [20,21,22].

Iohexol is a readily available compound, which can be used to measure the GFR. Its advantages include low cost, excellent safety profile, low protein binding, low levels of toxicity at the dose used for measuring GFR, stability at room temperature (20 to 25 °C), and being able to provide an accurate measure of glomerular filtration [20, 23]. Previous studies using iohexol to measure GFR in sub-Saharan Africa have included few people with CKD [13, 24, 25]. The largest study to date conducted in the Democratic Republic of Congo and Ivory Coast used plasma iohexol to measure GFR in 494 participants. They noted that the African-American race coefficient did not improve the performance of creatinine-based GFR estimates of iohexol GFR. In particular, the Chronic Kidney Disease Epidemiology (CKD-EPI) and Modification of Diet in Renal Disease (MDRD) equations performed better without the race coefficient in participants with GFR ≥60 mL/min/1.73m2. The authors also evaluated the Full Age Spectrum (FAS) equation and found it to be as accurate as CKD-EPI for GFR ≥60 mL/min/1.73m2 but better for those with creatinine based GFR < 60 mL/min/1.73m2. Addition of cystatin C did not improve performance of the equations among this study group [25].

In the proposed study, we will use iohexol excretion in a population sample of 3000 participants with and without CKD to measure GFR and determine the optimal equation to estimate GFR in Uganda, Malawi and South Africa.

Methods

Primary objectives

  1. 1.

    To measure GFR using plasma clearance of iohexol

  2. 2.

    To assess the performance of existing eGFR equations, namely Cockcroft -Gault (CG), Modification of Diet in Renal Disease (4-v MDRD) and Chronic Kidney Disease-Epidemiology collaboration (CKD-EPI) by comparing these to measured GFR, among sub-Saharan Africans.

Secondary objectives

In a Ugandan sub-sample, to assess the feasibility of using dry blood spot collection of iohexol in the determination of GFR.

In a Ugandan sub-sample to explore the current community understanding of kidney disease and available treatments from traditional healers and herbalists.

Study organization and sites

We will have three participant countries across SSA (Fig. 1) and these centres have formed a new collaborative group, African Research on Kidney Disease (ARK) around this initial study.

Fig. 1
figure1

Map of Africa showing the ARK centres from Malawi, South Africa and Uganda. The general population cohort is run by the Medical Research Council-Uganda virus Institute and the London School of Hygiene and tropical Medicine (MRC-UVRI-LSHTM, Uganda). The Malawi Epidemiology and Intervention Research Unit (MEIRU) runs the MEIRU cohort in Malawi and the Medical Research Council/Wits University Rural Public Health and Health Transitions Research Unit runs the Agincourt Unit in South Africa. Map proposed by Robert Kalyesubula and illustrated by Helmut Kraus

A working committee will review the protocols for the design, conduct, progress and data collection and analysis plan. The committee will meet in person on several occasions to review protocols and measurement tools and to agree on a minimum dataset. The three centres have unique characteristics that will enrich the study with diversity among participants. The details of each of the different cohorts in which the current study will be nested are described below:

Uganda

In Uganda, the study will be based in the General Population Cohort (GPC), originally established in 1989 [26]. The study area is located in rural south-western Uganda in one sub-county of Kalungu district, approximately 120 km from Entebbe town. The cohort comprises all residents of 25 adjacent villages, approximately 22,000 people with men and women in equal proportions, and 52% older than 13 years. The study population for the general population cohort is recruited through annual house-to-house rounds of the census where study participants are selected. Medical care is available through an established general clinic.

Malawi

The Malawi Epidemiology and Intervention Research Unit (MEIRU) conducts ongoing epidemiological studies based at two sites, which will be used for recruiting participants for this study. The first is Karonga District in rural northern Malawi, predominantly subsistence farmers, fishermen and informal traders. The second site is Area 25 of Lilongwe, the capital city of Malawi. The urban area is socio-economically mixed and includes government and industry employees, traders and those in casual employment. We will enrol participants from both sites, providing a true population-wide sample and enabling urban: rural comparisons.

The stratified study sample was drawn from a pool of over 5000 individuals who were tested for serum creatinine as part of a larger survey (n = 33,000) of chronic conditions. Medical care is available through an established chronic care Non-Communicable Disease (NCD) clinic.

South Africa

In South Africa, the study will be located within the Medical Research Council/University of the Witwatersrand Rural Public Health and Health Transitions Research Unit in Agincourt, Bushbuckridge sub-district of Mpumalanga province. The area includes a high-functioning health and socio-demographic surveillance system (HDSS) which covers approximately 115,000 people based on an annual update of resident status and vital events. The study setting also serves as a national pilot site for the development of integrated chronic disease care. The population still has gaps in access to electricity, water and tarred roads and unemployment rates are high, leading to high rates of labour migration. Patients will be managed through the primary health care system consisting of six clinics, two health centres and three district hospitals. The HDSS has a central clinic with advanced laboratory capacity, which will be used for this study.

Study design

This will consist of cross-sectional studies embedded within ongoing cohorts in Malawi and Uganda and a longitudinal study in South Africa, all with serial bio-banking.

Data collection, sample size and selection of sample

We intend to recruit 3000 participants in total. We have based the power calculation is based on the number of study participants needed to examine the accuracy of GFR estimating equations for ‘true’ GFR, at each of the three study sites separately. We have specified this as having 90% power to detect whether eGFR is within 5% of the iohexol value at an eGFR of 60mls/min, assuming a standard deviation of 25mls/min [9] and with a two-sided alpha of 0.05. This gives an estimated required sample size of 730 participants. Allowing for participants who do not wish to participate in this element of the study, and technical failures, we intend to recruit 1000 participants in each country.

In order to fulfil the aims of the study, a structured approach is needed. This will differ between South Africa and the other two sites (Malawi and Uganda).

Details of site recruitment

We plan to recruit from each of the sites guided by the previous recruitment protocols. Each of the sites will however, have two phases of recruitment.

For the Uganda GPC, the first phase will consist of measurement of creatinine from stored sera.

Participant selection for this part of the study has been described elsewhere [27]. Briefly, participants were selected from the general population cohort, which is a community-based open cohort study of residents of 25 neighbouring villages. The participants are selected through house-to-house mobilisation and community surveys conducted through village-based hubs. For Uganda, 5979 individuals were identified for baseline creatinine measurement, which included all the adults surveyed in that round. We will use creatinine levels from the stored samples to stratify participants for the second phase of the study. Here, we will select participants according to eGFR cut offs of normal (>90mls/min/1.73m2); impaired renal function (60-90mls/min/1.73m2) and low GFR (<60mls/min/1.73m2), all additionally stratified by age and sex. Selected participants will be approached by the community engagement lead (community mobilizer) for each village. Once participants assemble, the study team will take time to explain the details of the study to them, after which individual informed consent will be obtained. Participants will undertake a questionnaire to collect demographic data as well as history of known risk factors for kidney disease. Participants will also undergo detailed biophysical measurements and blood draws as detailed in the next sections. We will give participants an appointment to attend the medical clinic at the central Research Station, where iohexol measurements and other study tests will be performed.

For Malawi, for the first phase of the study, we will use creatinine assays from a previous survey to guide the selection of participants for the second phase of the study using a similar approach to that of Uganda. However, we will recruit participants 18 years and above from Malawi through household visits where similar procedures to those outlined above will be undertaken. We will carry out iohexol testing and other physical tests at a research clinic (phase 2). Details of participant recruitment have been detailed elsewhere [28].

In South Africa, we will determine baseline creatinine and the prevalence of CKD using an age and gender stratified random sample of 2250 members of the rural Agincourt HDSS. The number of participants sampled in each stratum will be determined in order to ensure proportional allocation, based on the population demographic distribution. We will visit participants in their homes for the first phase of the study where the questionnaires, biophysical measurements and take blood draws to determine Human immunodeficiency virus (HIV) status, blood sugar level, lipid profile as well as the baseline creatinine level. We will repeat creatinine and urine albumin measures after a minimum period of 3 months to confirm CKD. For the second phase of the study, we will recruit participants from Phase 1 through household visits and give them a referral date to the clinic for iohexol measurement along with other tests (See Table 1 and Fig. 2 for details).

Table 1 Clinical and Laboratory measurements within the ARK study, by country
Fig. 2
figure2

Recruitment flowchart for ARK. ABPM- Ambulatory Blood Pressure Monitor; ACR- Albumin: creatinine Ratio; ASOT- Antistreptococcal antibody titres; BIA- Bioimpedance Analysis; BP-Blood Pressure; CBC- Cell blood count; GFR- glomerular filtration rate; CKD- Chronic kidney disease; HB-Haemoglobin; NCD-Non-communicable diseases; SA-South Africa. Credit for the ARK study map and copyright go to Helmut Kraus

Inclusion criteria

  • Adults aged 18 years and above from the three population cohorts

  • Able to give informed consent

Exclusion criteria

  • Blood pressure > 180/110mmhg

  • Pregnancy

  • Breast-feeding mothers

  • Known allergy to iodine−containing substances

  • Uncontrolled seizures (defined as a seizure within the last 12 months).

  • Acute febrile illness

Demographic factors

We will collect data on demographic factors including age, sex, place of residence, education, occupation and livelihood, tobacco use, alcohol use, dietary history, physical activity. We will also take a medical history and treatment for chronic diseases including HIV, diabetes mellitus, hypertension, heart disease and CKD as well as previous and current use of traditional medicine and drugs.

Physical examination

We will measure height, weight, waist circumference and hip circumference and calculate the body mass index (BMI) and the waist hip ratio accordingly. We will classify BMI according to WHO categories (weight/height2: kg/m2): underweight (< 18.5 kg/m2), normal weight (18.5–24.99 kg/m2), overweight (25.0–29.99 kg/m2) and obese (> 30.0 kg/m2).

We will undertake cardiovascular assessment through blood pressure and 24-h ambulatory blood pressure (BP) measurements (ABPM) on a sub-sample of participants (Malawi and Uganda), and electrocardiography (ECG). We will measure BP using Omron® M6 (for small, medium and large participants) and Omron HBP 110 machines (for obese participants). We will measure BP in triplicate after at least 5 min of rest and take the mean of the last two readings as the true blood pressure. We will derive BP classification from the National Institute of Health guidelines: where participants with a systolic BP greater than 120 mmHg but less than 140 mmHg, and/or a diastolic BP greater than 80 mmHg but less than 90 mmHg will be classified as pre-hypertensive. We defined hypertension as having a diastolic BP greater than or equal to 90 mmHg, systolic BP greater than or equal to 140 mmHg or being on treatment for high blood pressure. 24-h ambulatory blood pressure will be undertaken on a selected number of participants with no hypertension, pre-hypertension and hypertension across the spectrum of eGFR ranges and will capture wake and sleep periods. We will use the ECG for assessment of LVH using the Sokolow-Lyon criteria [29].

We will perform bioimpedance analysis (BIA) using the Bodystat® machine to measure body fat in relation to lean body mass for parameters outlined in Table 1. We will also perform a Dual-energy X-ray absorptiometry (DXA) scan for body composition in South Africa and use it to examine validity of the BIA measurements across countries.

Laboratory investigations

The key laboratory measurement for this study will be iohexol (Omnipaque 300 and 350 mg I/mL, GE Healthcare) clearance. The iohexol measured GFR will serve as the gold standard for comparison with other methods.

Using aseptic technique study nurses will insert an IV line into the non-dominant arm of the participant and use the line for subsequent blood draws. We will administer a single dose of 5 ml of iohexol over 30 s through an intravenous cannula inserted in the dominant arm contralateral to the established IV line, followed by a 10 ml normal saline flush. We will weigh the iohexol dose in grams to a specificity of 0.01 g by weighing the syringe before and after administration of the iohexol. The research nurses will draw four milliliters of blood for the iohexol assay again at approximately 5, 120, 180 and 240 min after injection of iohexol, and record the exact time. We will calculate the GFR using a single compartment model based on iohexol clearance between 120 and 240 min.

In Uganda at the time points of 120 and 240 min, we will also collect a capillary sample of dried blood spots to compare to intravenous sampling of iohexol GFR as a modification from previous studies [11, 30]. Should this method prove to be accurate, it will greatly facilitate further studies in resource-limited settings. In order to avoid inter-laboratory variations, all plasma and blood spot samples of iohexol will be measured at the National Health Laboratory, Johannesburg, South Africa by ultra-pressure liquid chromatography-tandem mass spectrometry with identification and quantitation based on MRM transitions. The laboratory is accredited for the International Organization for Standardization (ISO) standard 15,189 and also participates in the internal quality control run at a high and a low level every 20 samples. The laboratory has participated in Equalis external control since 2017.

We will measure creatinine using the enzymatic method in Uganda and Jaffe for South Africa and Malawi standardized against an isotope dilution mass spectrometry (IDMS) method across all sites. We will adopt the recommended estimation and reporting format or measured renal function proposed by Fabian et al [31]. We will measure serum cystatin C to determine the accuracy of the cystatin-based CKD-EPI equations. We will measure Cystatin C using the standardized Roche Gen2 assay, which has been standardized against certified reference material (ERM-DA471/IFCC). To minimise batch variation, samples will be analysed at the end of the study at one central laboratory at MRC/UCRI&LSHTM for Uganda and Malawi while the samples in South Africa will be analysed at the National Health Laboratory, Johannesburg, South Africa. To understand laboratory differences and enable calibration we will cross-validate and measure 50-paired samples for creatinine in Uganda, Malawi and South Africa, and cystatin between South Africa and Uganda. Additional tests include full blood count (including haemoglobin level), tests for screening of infections (malaria, streptococcal infections, Hepatitis B, Hepatitis C and HIV Infection), C-reactive protein, random blood glucose, HbA1c and the fasting lipid profile as markers of metabolic risk.

We will perform four urine tests. We will take an early morning spot urine sample for quantitative urine albumin: creatinine ratio, an early morning urine dipstick analysis to qualitatively assess haematuria, leucocyturia and proteinuria, and microscopy will be performed on the centrifuged urine to screen for urinary schistosomiasis within 1–2 h of collection. We will take a 24-h urine collection for urinary albumin excretion, sodium and 24-h creatinine clearance in a sample of 300 participants from Uganda during the second phase of the study.

Details of the equipment used for the measurements are outlined in Table 1 below.

Qualitative sub-study in Uganda

The qualitative sub-study in Uganda will have two components: a study of people with abnormal renal function, and an enquiry into medicinal product supply and use sourced from traditional healers and herbalists.

This sub-study of people with abnormal renal function will include up to 50 participants found to have CKD during the recruitment of the 1000 participants for the overarching study. In order to recruit a comparator group for the 50 CKD participants, a further 50 participants without CKD of a similar age living in a neighboring house to each case will be invited to take part. Data from both cases and ‘controls’ will be collected through two in-depth interviews. These selections were made on an estimated sample to reach saturation.

For the sub-study with traditional healers/herbalists, we will approach ten traditional healers and herbalists and invite them to participate in one or more interviews to discuss the common ailments they treat and the types of herbal and other preparations that they use in their practice.

Bio-banking

Serum, plasma and urine samples will be collected, processed and frozen and transported to the central laboratory for additional tests that cannot be conducted at the originating laboratory. All three centers have quality assurance systems for processing, sample transportation and sample storage management systems [26, 32, 33]. All sites have received ethical approval for testing of specimens for genetic studies and exploration of new markers of CKD.

Data management

Data management will follow local established procedures ensuring quality, confidentiality and proper use of abnormal results to guide patient care when needed. Electronic data capture allows automatic data checks such as double entry of numbers and range checks. We will enter anonymized data into compatible statistical databases from centres in all three countries to for sharing between all investigators.

Quality assurance

All three centres have quality assurance procedures that support the studies at different levels. Study staff are trained and certified according to national and international guidelines governing research. The ARK team holds regular meetings to ensure that study protocols and research standards are adhered to. We will follow standard sample storage protocols for the biorepository.

Statistical and qualitative analysis plan

We will tabulate baseline characteristics stratified by country. We will perform initial analyses in a sample of 1500 participants selected to represent all centres, ages, sex, and eGFR categories for modification of the equation, and allow for subsequent validation using the remaining data. We plan to recombine the training and test sets for fitting a final model. We are also open to alternatives to data-splitting such as re-sampling techniques (cross-validation and bootstrapping), that would allow us to develop the model on the whole dataset and still validate predictive accuracy. Evidence from previous derivation of eGFR equations suggests that errors are multiplicative, so we will do analysis on the log (ioxehol-GFR) scale. To determine the measured GFR (iohexol-GFR); we will calculate the slope from the 3 samples at 120, 180 and 240 min points using the exact time of collection turn these into a GFR normalised to BSA using standard methods [13, 34]. We will also put in the R-value of the fit, which will be used to exclude ones where the fit is particularly poor. In our primary analysis we will evaluate the performance of the CockCroft Gault, 4v-MDRD, CKD-EPIcr, CKD-EPIcyst, CKD-EPIcr/cyst equations with and without ethnic correction factor by calculating the bias, precision and accuracy at 10% (P10) and 30% (P30) compared to iohexol GFR as done by previous studies [13]. Derivation of GFR estimating equations is rapidly developing and we will also consider evaluation against newly developed methods including the full age spectrum (FAS) and the Lund-Malmo GFR estimating equations [35, 36]. We will assess the performance of the different equations for eGFRs with and without ethnic factors using Lin’s Concordance Correlation Coefficient and Spearman correlation coefficients to evaluate the degree to which pairs of observations fall on the 45° line through the origin. We will calculate bias and relative bias. We will evaluate precision by the standard deviation of the bias (random error) and root-mean-square error. We will test the difference in P30 accuracy between eGFRs with the exact McNemar test.

We will use Bland-Altman plots to investigate the measurement error of existing equations of log(MDRD-eGFR), log(CKD-EPI-eGFR), log(CG-eGFR) and log(cystatin C-eGFR) compared to log(ioxehol-GFR). If needed, to develop modified correction factors for the setting, we will carry out regressions to predict log(ioxehol-GFR) as a function of age, gender, and creatinine/cystatin C. Predicted eGFRs will subsequently be compared with measured GFRs in the validation sample. We will investigate what happens if information on weight, height, and bioimpedance measures are added to the regression model and use computed R2 values to investigate whether anthropometry adds relevant information over and above existing equations.

All statistical analyses will be performed using STATA 15 SE (Stata Corp, Texas, USA).

For the qualitative study, we will record interviews with permission from the participant and transcribe and translate them. We will revise the interview guidance after the initial interviews in order to collect greater depth of data on emerging themes. We will review transcripts for accuracy and enter these into Dedoose, qualitative data analysis software. We will conduct data analysis using an iterative coding process, during the interview period. After data collection, we will perform open coding, and move to more refined codes, subthemes, and then themes. We will analyze themes and code them using a constant comparison method. We will apply the codes to interview transcripts and summary statements with representative quotes developed for each theme.

Ethical considerations

We have obtained ethical approvals from the Uganda Virus Research Institute, Research Ethics Committee (UVRI-REC-#HS 1978) the Uganda National Council for Science and Technology (UNCST-#SS 4283), from the Malawi National Health Sciences Research Committee (#1072) and the University of Witwatersrand Human Research Ethics Committee (#M160938).

We will obtain written informed consent from participants for the collection and analysis of genetic samples as well as iohexol testing and for the use of their clinical records for research purposes. We will seek approval for all study procedures including material transfer agreements from the respective ethical review boards. In case we diagnose medical conditions through study screening, we will refer the participants to appropriate medical facilities. We will refer participants found to have advanced CKD to appropriate nephrology services as directed by the study physicians in each centre.

Discussion

This study aims to establish the optimal approach to estimate GFR in sub-Saharan Africa. Our study has the advantage of drawing participants from three countries and both rural and urban settings within sub-Saharan Africa, which will increase the applicability of the findings across the region. Furthermore, we embedded the ARK study within established cohorts that have background data and longitudinal serial measures that we can use to characterize kidney disease over a period. The study will seek to overcome a number of the limitations of previous research.

There are challenges with measuring iohexol GFR, with the major one being accuracy of measurement of iohexol [23, 37]. Stringent measures, shared standard operating procedures across the centres, and analysis in one certified laboratory will help minimise errors.

Measurement of creatinine is another challenge. Creatinine is influenced by many factors and these can greatly influence the estimation, particularly at low eGFR [19, 37, 38]. It is particularly important that the methods used for creatinine measurements are the same across sites and are validated across the centres. We opted to use the Jaffe method for South Africa and Malawi and enzymatic method in Uganda, standardized against an isotope dilution mass spectrometry method across all sites to improve the accuracy in measurement. We will measure Cystatin C in order to assess whether the addition of cystatin C by itself or in combination with creatinine assays further refines the accuracy of estimating GFR within SSA. Several studies have shown that the addition of cystatin C to the estimation formula improves its accuracy [39, 40] and using CKD-EPI-creatinine/cystatin C improved the accuracy in a study from Congo [13]. Using a unified approach, these issues will be addressed by this study. We also plan to measure bioimpedance to define body composition and its contribution to measured GFR variation across the sites, and to assess the correlation of body composition measured by BIA with the gold standard, DXA – which will be done in South Africa [41]. Bioimpedance measurement was included in this study for measurement of lean mass and total body water and we hope to use it to optimize GFR estimation on an individual patient basis, for example in a patient with significant oedema. While this has not proven useful in high-income settings, such technologies may be useful in low-resource settings [42]. Bio-impedance will help in incorporating more parameters allowing more flexibility to work with the data to get the best formula.

Assessment of risk factors for CKD in sub-Saharan Africa has previously been hampered by inaccurate measurement of GFR. Although, not exhaustive, we included a number of risk factors for CKD in our study. We will examine the role of traditional cardiovascular risk factors including diabetes mellitus, obesity and lipid profiles along with conventional and ambulatory blood pressure monitoring. In addition we will measure the impact of infections which may be of great importance in this region where infectious diseases are common and may play a major role in disease causation alongside genetic factors [43,44,45,46,47,48,49].

However, there are limitations. Prolonged blood sampling for iohexol excretion may help to define GFR in patients with very poor renal function [23]. We chose not to include this in our protocol to minimize participant burden. In addition, we do not include participants from other regions of sub-Saharan Africa, including Central and West Africa but we welcome data-sharing with studies in other regions.

The African Research into Kidney disease (ARK) collaboration will provide definitive information about optimal measurement of renal function in sub-Saharan Africa, and is an opportunity for establishing close working relationships across different centres, using standardized protocols and measurements, and shared bio-repositories. We plan to build on the collaboration for this study and for future work on kidney disease in sub-Saharan Africa, and welcome additional partners from across the region including North African and Arab populations.

Availability of data and materials

Not applicable

Abbreviations

ARK:

African Research into Kidney Diseases

CKD:

Chronic kidney disease

CKD-EPI:

Chronic Kidney Disease-Epidemiology collaboration

DTPA:

Technetium-99 m diethylenetriamine penta-acetic acid

EDTA:

chromium-51 ethylene diamine tetra-acetic acid

EGFR:

Estimated glomerular filtration rate

IDMS:

Isotope dilution mass spectrometry

MDRD:

Modification of Diet in Renal Disease

MGFR:

Measured glomerular filtration rate

References

  1. 1.

    Jha V, Garcia-Garcia G, Iseki K, Li Z, Naicker S, Plattner B, Saran R, Wang AY, Yang CW. Chronic kidney disease: global dimension and perspectives. Lancet. 2013;382(9888):260–72.

    PubMed  Article  PubMed Central  Google Scholar 

  2. 2.

    Stanifer JW, Jing B, Tolan S, Helmke N, Mukerjee R, Naicker S, Patel U. The epidemiology of chronic kidney disease in sub-Saharan Africa: a systematic review and meta-analysis. Lancet Glob Health. 2014;2(3):e174–81.

    PubMed  Article  PubMed Central  Google Scholar 

  3. 3.

    Ji M, Lee YH, Hur M, Kim H, Cho HI, Yang HS, Navarin S, Di Somma S, Network G. Comparing results of five glomerular filtration rate-estimating equations in the Korean general population: MDRD study, revised Lund-Malmo, and three CKD-EPI equations. Ann Lab Med. 2016;36(6):521–8.

    PubMed  PubMed Central  Article  Google Scholar 

  4. 4.

    Feng JF, Qiu L, Zhang L, Li XM, Yang YW, Zeng P, Guo XZ, Qin Y, Liu HC, Han XM, et al. Multicenter study of creatinine- and/or cystatin C-based equations for estimation of glomerular filtration rates in Chinese patients with chronic kidney disease. PLoS One. 2013;8(3):e57240.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  5. 5.

    Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16(1):31–41.

    CAS  Article  Google Scholar 

  6. 6.

    Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of diet in renal disease study group. Ann Intern Med. 1999;130(6):461–70.

    CAS  PubMed  Article  Google Scholar 

  7. 7.

    Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150(9):604–12.

    PubMed  PubMed Central  Article  Google Scholar 

  8. 8.

    Shemesh O, Golbetz H, Kriss JP, Myers BD. Limitations of creatinine as a filtration marker in glomerulopathic patients. Kidney Int. 1985;28(5):830–8.

    CAS  PubMed  Article  Google Scholar 

  9. 9.

    Delanaye P, Mariat C, Maillard N, Krzesinski JM, Cavalier E. Are the creatinine-based equations accurate to estimate glomerular filtration rate in African American populations? Clin J Am Soc Nephrol. 2011;6(4):906–12.

    PubMed  Article  Google Scholar 

  10. 10.

    van Deventer HE, George JA, Paiker JE, Becker PJ, Katz IJ. Estimating glomerular filtration rate in black south Africans by use of the modification of diet in renal disease and Cockcroft-Gault equations. Clin Chem. 2008;54(7):1197–202.

    PubMed  Article  CAS  Google Scholar 

  11. 11.

    Wyatt CM, Schwartz GJ, Owino Ong'or W, Abuya J, Abraham AG, Mboku C, M’Mene LB, Koima WJ, Hotta M, Maier P, et al. Estimating kidney function in HIV-infected adults in Kenya: comparison to a direct measure of glomerular filtration rate by iohexol clearance. PLoS One. 2013;8(8):e69601.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  12. 12.

    Madala ND, Nkwanyana N, Dubula T, Naiker IP. Predictive performance of eGFR equations in south Africans of African and Indian ancestry compared with (9)(9)mTc-DTPA imaging. Int Urol Nephrol. 2012;44(3):847–55.

    PubMed  Article  Google Scholar 

  13. 13.

    Bukabau JB, Sumaili EK, Cavalier E, Pottel H, Kifakiou B, Nkodila A, Makulo JRR, Mokoli VM, Zinga CV, Longo AL, et al. Performance of glomerular filtration rate estimation equations in Congolese healthy adults: the inopportunity of the ethnic correction. PLoS One. 2018;13(3):e0193384.

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  14. 14.

    Kaze FF, Halle MP, Mopa HT, Ashuntantang G, Fouda H, Ngogang J, Kengne AP. Prevalence and risk factors of chronic kidney disease in urban adult Cameroonians according to three common estimators of the glomerular filtration rate: a cross-sectional study. BMC Nephrol. 2015;16:96.

    PubMed  PubMed Central  Article  Google Scholar 

  15. 15.

    Myers GL, Miller WG, Coresh J, Fleming J, Greenberg N, Greene T, Hostetter T, Levey AS, Panteghini M, Welch M, et al. Recommendations for improving serum creatinine measurement: a report from the laboratory working Group of the National Kidney Disease Education Program. Clin Chem. 2006;52(1):5–18.

    CAS  PubMed  Article  Google Scholar 

  16. 16.

    Stevens LA, Manzi J, Levey AS, Chen J, Deysher AE, Greene T, Poggio ED, Schmid CH, Steffes MW, Zhang YL, et al. Impact of creatinine calibration on performance of GFR estimating equations in a pooled individual patient database. Am J Kidney Dis. 2007;50(1):21–35.

    CAS  PubMed  Article  Google Scholar 

  17. 17.

    Murthy K, Stevens LA, Stark PC, Levey AS. Variation in the serum creatinine assay calibration: a practical application to glomerular filtration rate estimation. Kidney Int. 2005;68(4):1884–7.

    CAS  PubMed  Article  Google Scholar 

  18. 18.

    Verhave JC, Balje-Volkers CP, Hillege HL, de Zeeuw D, de Jong PE. The reliability of different formulae to predict creatinine clearance. J Intern Med. 2003;253(5):563–73.

    CAS  PubMed  Article  Google Scholar 

  19. 19.

    Coresh J, Astor BC, McQuillan G, Kusek J, Greene T, Van Lente F, Levey AS. Calibration and random variation of the serum creatinine assay as critical elements of using equations to estimate glomerular filtration rate. Am J Kidney Dis. 2002;39(5):920–9.

    CAS  PubMed  Article  Google Scholar 

  20. 20.

    Levey AS, Inker LA. Assessment of glomerular filtration rate in health and disease: a state of the art review. Clin Pharmacol Ther. 2017;102(3):405–19.

    CAS  PubMed  Article  Google Scholar 

  21. 21.

    George C, Mogueo A, Okpechi I, Echouffo-Tcheugui JB, Kengne AP. Chronic kidney disease in low-income to middle-income countries: the case for increased screening. BMJ Glob Health. 2017;2(2):e000256.

    PubMed  PubMed Central  Article  Google Scholar 

  22. 22.

    Vega JA, Ochola PS, Marsh ML, HO VQ, Fisher T. Comparison of 24-hourour urine to estimated renal function using CKD-EPI, MDRD4 and cockcroft-gault in specific patient subsets. J Pharma Care Health Sys. 2015;2(129):2376.

    Google Scholar 

  23. 23.

    Delanaye P, Ebert N, Melsom T, Gaspari F, Mariat C, Cavalier E, Bjork J, Christensson A, Nyman U, Porrini E, et al. Iohexol plasma clearance for measuring glomerular filtration rate in clinical practice and research: a review. Part 1: how to measure glomerular filtration rate with iohexol? Clin Kidney J. 2016;9(5):682–99.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  24. 24.

    Sagou Yayo E, Aye M, Konan JL, Emieme A, Attoungbre ML, Gnionsahe A, Cavalier E, Monnet D, Delanaye P. Inadequacy of the African-American ethnic factor to estimate glomerular filtration rate in an African general population: results from cote d’Ivoire. Nephrol Ther. 2016;12(6):454–9.

    PubMed  Article  Google Scholar 

  25. 25.

    Bukabau JB, Yayo E, Gnionsahe A, Monnet D, Pottel H, Cavalier E, Nkodila A, Makulo JRR, Mokoli VM, Lepira FB, et al. Performance of creatinine- or cystatin C-based equations to estimate glomerular filtration rate in sub-Saharan African populations. Kidney Int. 2019;95(5):1181–9.

    CAS  PubMed  Article  Google Scholar 

  26. 26.

    Asiki G, Murphy G, Nakiyingi-Miiro J, Seeley J, Nsubuga RN, Karabarinde A, Waswa L, Biraro S, Kasamba I, Pomilla C, et al. The general population cohort in rural South-Western Uganda: a platform for communicable and non-communicable disease studies. Int J Epidemiol. 2013;42(1):129–41.

    PubMed  PubMed Central  Article  Google Scholar 

  27. 27.

    Kalyesubula R, Hau JP, Asiki G, Ssebunya B, Kusemererwa S, Seeley J, Smeeth L, Tomlinson L, Newton R. Impaired renal function in a rural Ugandan population cohort. Wellcome Open Res. 2018;3:149.

    PubMed  Article  Google Scholar 

  28. 28.

    Nakanga WP, Prynn JE, Banda L, Kalyesubula R, Tomlinson LA, Nyirenda M, Crampin AC. Prevalence of impaired renal function among rural and urban populations: findings of a cross-sectional study in Malawi. Wellcome Open Res. 2019;4:92.

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  29. 29.

    Hancock EW, Deal BJ, Mirvis DM, Okin P, Kligfield P, Gettes LS, Bailey JJ, Childers R, Gorgels A, Josephson M, et al. AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part V: electrocardiogram changes associated with cardiac chamber hypertrophy: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society. Endorsed by the International Society for Computerized Electrocardiology. J Am Coll Cardiol. 2009;53(11):992–1002.

    PubMed  Article  Google Scholar 

  30. 30.

    Niculescu-Duvaz I, D'Mello L, Maan Z, Barron JL, Newman DJ, Dockrell ME, Kwan JT. Development of an outpatient finger-prick glomerular filtration rate procedure suitable for epidemiological studies. Kidney Int. 2006;69(7):1272–5.

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  31. 31.

    Fabian J, George JA, Etheredge HR, Deventer M, Kalyesubula R, Wade AN, Tomlinson LA, Tollman S, Naicker S, African Research in Kidney Disease ARK. Methods and reporting of kidney function: a systematic review of studies from sub-Saharan Africa. Clin Kidney J. 2019;12:778.

    PubMed  PubMed Central  Article  Google Scholar 

  32. 32.

    Crampin AC, Kayuni N, Amberbir A, Musicha C, Koole O, Tafatatha T, Branson K, Saul J, Mwaiyeghele E, Nkhwazi L, et al. Hypertension and diabetes in Africa: design and implementation of a large population-based study of burden and risk factors in rural and urban Malawi. Emerg Themes Epidemiol. 2016;13:3.

    PubMed  PubMed Central  Article  Google Scholar 

  33. 33.

    Kahn K, Collinson MA, Gomez-Olive FX, Mokoena O, Twine R, Mee P, Afolabi SA, Clark BD, Kabudula CW, Khosa A, et al. Profile: Agincourt health and socio-demographic surveillance system. Int J Epidemiol. 2012;41(4):988–1001.

    PubMed  PubMed Central  Article  Google Scholar 

  34. 34.

    Peters AM. Expressing glomerular filtration rate in terms of extracellular fluid volume. Nephrol Dial Transplant. 1992;7(3):205–10.

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  35. 35.

    Delanaye P, Ebert N, Pottel H. Newer GFR estimating equations require validation in different populations. Am J Kidney Dis. 2017;70(4):586.

    PubMed  Article  PubMed Central  Google Scholar 

  36. 36.

    Bjork J, Nyman U, Berg U, Delanaye P, Dubourg L, Goffin K, Grubb A, Hansson M, Littmann K, Asling-Monemi K, et al. Validation of standardized creatinine and cystatin C GFR estimating equations in a large multicentre European cohort of children. Pediatr Nephrol. 2019;34(6):1087–98.

    PubMed  Article  PubMed Central  Google Scholar 

  37. 37.

    Seegmiller JC, Eckfeldt JH, Lieske JC. Challenges in measuring glomerular filtration rate: a clinical laboratory perspective. Adv Chronic Kidney Dis. 2018;25(1):84–92.

    PubMed  Article  PubMed Central  Google Scholar 

  38. 38.

    Seronie-Vivien S, Galteau MM, Carlier MC, Hadj-Aissa A, Hanser AM, Hym B, Marchal A, Michotey O, Pouteil-Noble C, Sternberg M, et al. Improving the interlaboratory variation for creatinine serum assay. Ann Biol Clin. 2004;62(2):165–75.

    CAS  Google Scholar 

  39. 39.

    Werner K, Pihlsgard M, Elmstahl S, Legrand H, Nyman U, Christensson A. Combining cystatin C and creatinine yields a reliable glomerular filtration rate estimation in older adults in contrast to beta-trace protein and beta2-microglobulin. Nephron. 2017;137(1):29–37.

    CAS  PubMed  Article  Google Scholar 

  40. 40.

    Bjork J, Grubb A, Larsson A, Hansson LO, Flodin M, Sterner G, Lindstrom V, Nyman U. Accuracy of GFR estimating equations combining standardized cystatin C and creatinine assays: a cross-sectional study in Sweden. Clin Chem Lab Med. 2015;53(3):403–14.

    PubMed  Article  CAS  Google Scholar 

  41. 41.

    Ozmen S, Kaplan MA, Kaya H, Akin D, Danis R, Kizilkan B, Yazanel O. Role of lean body mass for estimation of glomerular filtration rate in patients with chronic kidney disease with various body mass indices. Scand J Urol Nephrol. 2009;43(2):171–6.

    CAS  PubMed  Article  Google Scholar 

  42. 42.

    Frank M, Guarino-Gubler S, Burnier M, Maillard M, Keller F, Gabutti L. Estimation of glomerular filtration rate in hospitalised patients: are we overestimating renal function? Swiss Med Wkly. 2012;142:w13708.

    PubMed  Google Scholar 

  43. 43.

    Ulasi II, Tzur S, Wasser WG, Shemer R, Kruzel E, Feigin E, Ijoma CK, Onodugo OD, Okoye JU, Arodiwe EB, et al. High population frequencies of APOL1 risk variants are associated with increased prevalence of non-diabetic chronic kidney disease in the Igbo people from South-Eastern Nigeria. Nephron Clin Pract. 2013;123(1–2):123–8.

    PubMed  Article  Google Scholar 

  44. 44.

    Stanifer JW, Karia F, Maro V, Kilonzo K, Qin X, Patel UD, Hauser ER. APOL1 risk alleles among individuals with CKD in northern Tanzania: a pilot study. PLoS One. 2017;12(7):e0181811.

    PubMed  PubMed Central  Article  Google Scholar 

  45. 45.

    Osafo C, Raji YR, Olanrewaju T, Mamven M, Arogundade F, Ajayi S, Ulasi I, Salako B, Plange-Rhule J, Mengistu Y, et al. Genomic approaches to the burden of kidney disease in sub-Saharan Africa: the human heredity and health in Africa (H3Africa) kidney disease research network. Kidney Int. 2016;90(1):2–5.

    PubMed  PubMed Central  Article  Google Scholar 

  46. 46.

    Kasembeli AN, Duarte R, Ramsay M, Mosiane P, Dickens C, Dix-Peek T, Limou S, Sezgin E, Nelson GW, Fogo AB, et al. APOL1 risk variants are strongly associated with HIV-associated nephropathy in black south Africans. J Am Soc Nephrol. 2015;26(11):2882–90.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  47. 47.

    Okunola O, Akinsola A, Ayodele O. Kidney diseases in Africa: aetiological considerations, peculiarities and burden. Afr J Med Med Sci. 2012;41(2):119–33.

    CAS  PubMed  Google Scholar 

  48. 48.

    Arogundade FA, Hassan MO, Omotoso BA, Oguntola SO, Okunola OO, Sanusi AA, Akinsola A. Spectrum of kidney diseases in Africa: malaria, schistosomiasis, sickle cell disease, and toxins. Clin Nephrol. 2016;86(13):53–60.

    PubMed  Article  Google Scholar 

  49. 49.

    Okpechi IG, Ameh OI, Bello AK, Ronco P, Swanepoel CR, Kengne AP. Epidemiology of histologically proven glomerulonephritis in Africa: a systematic review and meta-analysis. PLoS One. 2016;11(3):e0152203.

    PubMed  PubMed Central  Article  CAS  Google Scholar 

Download references

Acknowledgements

We would like to thank all the participants for the primary studies as well as the staff for the ARK cohorts. We will like to thank Professor Anatoli Kamali for his involvement in the initial stages of setting up the study.

Funding

The study in Uganda and Malawi (RK) is funded by the GlaxoSmithKline Africa Non-Communicable Disease (NCD) Open Lab (Project Number: 8111). The study in South Africa (SN) is funded by the South African Medical Research Council, MRC UK (via the Newton Fund) and the GlaxoSmithKline Africa NCD Open Lab (Project Number: 074). LAT is funded by a Wellcome Trust Intermediate Clinical Fellowship (101143/Z/13/Z). The original cohort studies at MEIRU, Malawi were funded by the Wellcome Trust. ANW is supported by the Fogarty International Centre (NIH) under Award Number K43TW010698. This paper describes the views of the authors and does not necessarily represent the official views of the National Institutes of Health (USA).

The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.

Author information

Affiliations

Authors

Contributions

RK, JN, RN, DN, MN, LS, SN, ACC and LAT conceived the study, designed the study, drafted work and substantially revised the manuscript. CH, WN, BS, JP, JG, ANW, JS designed the study, drafted work and substantially revised the manuscript. All authors have approved submission of the manuscript and subscribe to the contents as accurate.

Corresponding author

Correspondence to Robert Kalyesubula.

Ethics declarations

Ethics approval and consent to participate

We have obtained ethical approvals from the Uganda Virus Research Institute, Research Ethics Committee (UVRI-REC) the Uganda National Council for Science and Technology (UNCST), from the Malawi National Health Sciences Research Committee and the University of Witwatersrand Human Research Ethics Committee (Medical).

We will obtain written informed consent from participants for the collection and analysis of genetic samples as well as iohexol testing and for the use of their clinical records for research purposes. We will seek approval for all study procedures including material transfer agreements from the respective ethical review boards. In case we diagnose medical conditions through study screening, we will refer the participants to appropriate medical facilities. We will refer participants found to have advanced CKD to appropriate nephrology services as directed by the study physicians in each centre.

Consent for publication

Not applicable

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

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

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kalyesubula, R., Fabian, J., Nakanga, W. et al. How to estimate glomerular filtration rate in sub-Saharan Africa: design and methods of the African Research into Kidney Diseases (ARK) study. BMC Nephrol 21, 20 (2020). https://doi.org/10.1186/s12882-020-1688-0

Download citation

Keywords

  • Measurement
  • Glomerular filtration rate
  • Chronic kidney disease
  • Sub-Saharan Africa