PIIS2214109X14700026
Articles The epidemiology of chronic kidney disease in sub-Saharan
Africa: a systematic review and meta-analysis
John W Stanifer, Bocheng Jing, Scott Tolan, Nicole Helmke, Romita Mukerjee, Saraladevi Naicker, Uptal Patel
Summary
Background Amid rapid urbanisation, the HIV epidemic, and increasing rates of non-communicable diseases, people
in sub-Saharan Africa are especially vulnerable to kidney disease. Little is known about the epidemiology of chronic
kidney disease (CKD) in sub-Saharan Africa, so we did a systematic review and meta-analysis examining the epidemiology of the disease.
Methods We searched Medline, Embase, and WHO Global Health Library databases for all articles published through
March 29, 2012, and searched the reference lists of retrieved articles. We independently reviewed each study for
quality. We used the inverse-variance random-e? ects method for meta-analyses of the medium-quality and high-
quality data and explored heterogeneity by comparing CKD burdens across countries, settings (urban or rural),
comorbid disorders (hypertension, diabetes, HIV), CKD de? nitions, and time.
Findings Overall, we included 90 studies from 96 sites in the review. Study quality was low, with only 18 (20%)
medium-quality studies and three (3%) high-quality studies. We noted moderate heterogeneity between the medium-
quality and high-quality studies (n=21; I2=47·11%, p<0·0009). Measurement of urine protein was the most common
method of determining the presence of kidney disease (62 [69%] studies), but the Cockcroft-Gault formula (22 [24%]
studies) and Modi? cation of Diet in Renal Disease formula (17 [19%] studies) were also used. Most of the studies
were done in urban settings (83 [93%] studies) and after the year 2000 (57 [63%] studies), and we detected no
signi? cant di? erence in the prevalence of CKD between urban (12·4%, 95% CI 11–14) and rural (16·5%, 13·8–19·6)
settings (p=0·474). The overall prevalence of CKD from the 21 medium-quality and high-quality studies was 13·9%
(95% CI 12·2–15·7).
Interpretation In sub-Saharan Africa, CKD is a substantial health burden with risk factors that include communicable
and non-communicable diseases. However, poor data quality limits inferences and draws attention to the need for
more information and validated measures of kidney function especially in the context of the growing burden of non-communicable diseases.
Funding Duke University.
Copyright ? Stanifer et al. Open Access article distributed under the terms of CC BY-NC-ND
Introduction
Amid population ageing, lifestyle changes, and rapid urbanisation, the importance of non-communicable diseases in low-income and middle-income countries is becoming increasingly recognised, and in 2011 the United Nations General Assembly adopted a resolution acknowledging the impending global burden and threat of non-communicable diseases.1 However, few epi-demiological studies of the incidence, prevalence, and cause of these diseases have been done.1–3 Chronic kidney disease (CKD), which nearly doubled as a cause of death worldwide between 1990 and 2010 and was the 18th highest cause of death worldwide in 2010, falls into this category of diseases.4
Sub-Saharan Africa is a vast and heterogeneous region of roughly 24 million km2 composed of 47 countries and more than 900 million people.5 By 2030, more than 70% of patients with end-stage renal disease are estimated to be living in low-income countries, such as those in sub-Saharan Africa, where the gross domestic product per person is on average
less than US$1500 per year.6–8 This estimation is
alarming in view of the fact that the global prevalence
of maintenance dialysis has doubled since 1990, and
that renal replacement therapy was accessed by
1·8 million people worldwide in 2004 with less than 5%
of that population coming from sub-Saharan Africa.9,10
There are many potential causes of CKD in sub-
Saharan Africa, making kidney disease especially
burdensome in the region. In addition to non-
communicable diseases, communicable diseases
such as infectious glomerulonephritis, schistosomiasis,
leishmaniasis, and HIV infection are common and can
cause CKD. And because more than 22 million people in
sub-Saharan Africa have HIV, the potential for an
overwhelming burden of CKD in the region is high.11–14
As has been noted elsewhere,15,16 the challenging ? rst
step to address the burden of kidney disease in Africa is to
establish the epidemiology of CKD. We hypothesised that
CKD is an important disease in sub-Saharan Africa with
Lancet Glob Health 2014
Published Online
February 10, 2014
https://www.wendangku.net/doc/8716253427.html,/10.1016/
S2214-109X(14)70002-6
See Online/Comment
https://www.wendangku.net/doc/8716253427.html,/10.1016/
S2214-109X(14)70014-2
Department of Medicine
(J W Stanifer MD, S Tolan MD,
U Patel MD), Duke Global Health
Institute (J W Stanifer),
Department of Biostatistics
(B Jing BS),School of Medicine
(N Helmke BS), and Duke
Clinical Research Institute
(R Mukerjee MD, U Patel),Duke
University, Durham, NC, USA;
and Department of Internal
Medicine, University of
Witwatersrand, Johannesburg,
South Africa
(Prof S Naicker MBChB)
Correspondence to:
Dr John W Stanifer, Duke
University Medical Center,
Box 3182, Durham,
NC 27710, USA
john.stanifer@https://www.wendangku.net/doc/8716253427.html,
Articles
many potential causes. To assess the extent to which it is a public health and economic burden, we did a systematic review and meta-analysis to ?nd out the prevalence of CKD in adults in the region.
Methods
Search strategy and selection criteria
We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.17 We searched PubMed, Embase, and the WHO Global Health Library databases (which include the African [AF RO] Index Medicus, WHO Library Information System [WHOLIS], and Scienti? c Electronic Library Online [SciELO]) for published reports of kidney disease in sub-Saharan Africa up to March 29, 2012. We used Boolean logic with search terms including “chronic kidney disease”, “nephropathy”, “renal insu? ciency”, and “Africa south
of Sahara”. We used controlled vocabularies (eg, Medical Subject Heading terms) to identify synonyms. We applied no language restrictions. The study protocol and detailed search parameters are available in the appendix. Two authors (JWS and RM) independently reviewed each title and abstract for inclusion. Randomised controlled trials and non-randomised studies including cross-sectional studies, cohort studies (retrospective and prospective), case-control studies, and literature reviews that report a prevalence or a calculable prevalence for CKD in adults from a sub-Saharan African country were eligible for inclusion. We excluded studies that were only composed of pregnant women or people admitted to hospital, or ones that speci? cally recruited participants on the basis of the presence or absence of kidney disease. We regarded the following terms as equivalent to chronic kidney disease: “renal insu? ciency”, “chronic kidney failure”, “renal impairment”, “end/target-organ damage” (if renal function was recorded in some capacity), “end-stage kidney/renal disease”, “nephropathy”, “proteinuria”, “dialysis-dependent”, “status-post renal transplant”, “sonographic-evidence”, and “biopsy-proven renal disease”.
Two investigators (JWS and ST) did a second review assessing entire papers and their reference lists. Any disagreement resulted in joint review of the article with reconciliation. Articles in French were translated by one author (NH) and then independently reviewed by two authors (JWS and ST or JWS and RM) for inclusion and extraction (only one article was not in French or English, and it was excluded on the basis of title and abstract alone). Detailed inclusion and exclusion criteria are available in the appendix.
Quality assessment and data extraction
Three authors (JWS, ST, and RM) independently appraised each article for quality using previously described assessment criteria (panel).18–23 On the basis of these criteria, each article received a quality grade of low, medium, or high, and any disagreement resulted in joint review of the article.
Data were extracted from all included studies by two authors independently (JWS and ST) and in duplicate into a preformulated table; errors of data extraction were resolved by joint review of the original articles. In addition to the overall prevalence reported by each study, we extracted or calculated, as necessary, the prevalence of CKD as de? ned by proteinuria, a creatinine clearance of less than 60 mL per min by the Cockcroft-Gault formula,
See Online for appendix
Panel: Quality assessment criteria for studies examining the prevalence of chronic kidney disease
High quality
For studies of the highest quality, assessors should answer yes to the following ten questions
1 Subject sampling and precision
A Are the included people representative of the general population? (Comment:
if people were included on the basis of hospital records, insurance claims, or
health-care facilities then they should not be considered representative of the
general population.)
B People are not included or excluded on the basis of speci? c risk factors. (Comment:
high risk people such as those with diabetes, HIV, or hypertension should not be
sought out speci? cally for inclusion or exclusion.)
C Is the sam p le size adequate to address the question of prevalence in the studied
population?
2 Sampling technique
A Were the people recruited at random? (Comment: methods should address the
issue of enrolling consecutive participants, people likely to have the disease or at
high risk, and convenience sampling)
3 Response r ate
A Does the article report a response rate in total sample?
B Is that response rate 40% or higher?
4 Exclusion r ate
A Does the article report an exclusion rate in total sample?
B Is the exclusion rate 10% or less?
5 Measurement and method of determination of kidney disease
A Does the study report the method used for determination of kidney disease?
B Does the study use a consistent method for determination of kidney disease? Medium quality
For studies of medium quality, assessors should answer yes to the following questions
1 If participants are not representative of the general population, then are they
representative of the population in question?(Comment: people can be taken from high-risk groups such as diabetic or hypertensive populations; although not considered representative of the entire population, they can still be representative of that speci? c population.)
2 If participants were not recruited at random, then were they recruited in a random
non-health-care convenience method from the entire population in question?
3 Is the study sample size adequate to answer the question of prevalence in the studied
population?
4 Does the study use a consistent method for determination of kidney disease?
Low quality
1 For studies of the lowest quality, assessors would be unable to answer yes to all of the above
questions.
Articles
or an estimated glomerular ? ltration rate (GFR) of less than 60 mL per min per 1·73 m2 by the Modi? cation of Diet in Renal Disease (MDRD) and Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formulae. Although prevalence was the primary summary measure of interest, we also extracted the following information from each study: authors, location, setting, study population, dates, comorbidities, sample size, methods of measurement for kidney disease, mean age, and sex. If studies did not specify the exact years of study, we used the year of publication. We extracted or calculated these speci? c data only when we were able to con? dently establish that patients were not overlapping or being counted twice. In cases of uncertainty or missing data, we contacted the corresponding authors for additional information.
Statistical analysis
We investigated heterogeneity using an I2 statistic and completed meta-analyses with the inverse-variance random-e? ects method, as described by DerSimonian and Laird.24 Each study prevalence estimate received a weight that was equal to the reciprocal of within-study variance (v
i
) summed with between-study variance (τ2). The calculated tau-squared (τ2) or between-study variance for our analysis was 0·010766. Only medium-quality and high-quality studies were included in the meta-analyses. We then calculated pooled prevalence as a weighted mean among all the study estimates. We examined prevalence by region, country, method of kidney disease measurement, setting (urban or rural), date, and disease-speci? c populations (HIV, hypertension, and diabetes type 1 and 2).
We assessed inter-rater agreement for inclusion using Cohen’s kappa (κ) coe?cient, and assessed inter-rater agreement for quality assessment using F leiss’s κ statistic. We calculated 95% CIs using Wilson’s score method. This method for determination of two-sided CIs
of proportions (prevalence) prohibits overshoot of zero or one and is applicable to all sample sizes.25,26 We used a Welch two-sided t test and Wilcoxon rank sum test to compare prevalence estimates between groups. All continuous variables were reported as mean (SD). We used R (version 2.15.0; rmeta package)for all statistical analyses.
Role of the funding source
The sponsor of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. All authors had full access to all the data in the study and had ?nal responsibility for the decision to submit for publication.
Results
We identi? ed 905 articles, of which 90 were included in the systematic review (?gure 1; appendix). Inter-rater agreement for inclusion was excellent (κ=0·974), and any disagreement resulted in inclusion for full text review.The 90 articles, done between 1962 and 2011, reported
data for 21 countries with a geographical distribution
covering most of sub-Saharan Africa. However, more
than half the studies (48 [53%] studies) were from South
Africa, Nigeria, and Ethiopia. Data for 64 307 people (42%
men) were included, with a mean age of 41·4 years
(SD 9·9). 46 494 (72%) people had diabetes, 37 169 (58%)
had HIV, 7845 (12%) had hypertension, and 2765 (4%)
were obese (body-mass index [BMI]>30 kg/m2).
18 (20%) studies27–44 from 25 sites met criteria for the
medium level of quality, and three (3%) studies45–47 met
criteria for the high level of quality. Inter-rater agreement
for quality assessment was excellent (κ=0·916). The
common reasons for studies not meeting the criteria were
the sampling methods used, most commonly convenience
sampling, and unreliable measurements of kidney
function. Most studies (72 [80%] studies) examined
populations thought to be high risk—eg, people with HIV,
diabetes, hypertension, or other risk factors for kidney
disease. The studies assessed kidney function by urine
protein (62 [69%] studies), estimation of creatinine
clearance by Cockcroft-Gault formula (22 [24%] studies),
and estimation of GF R by MDRD formula (17 [19%]
studies). Less often, biopsy (two [2%] studies) and
calculation of GFR by CKD-EPI formula (two [2%] studies)
were also used. Other measurements of kidney function
included simple chart review, absolute creatinine value,
Figure 1: Flow diagram of study selection
* WHO Global Health Library databases include the African Index Medicus, Medline, World Health Organization Library Information System (WHOLIS), and Scienti? c Electronic Library Online (SciELO).
Articles
and presence of azotaemia as indicated by absolute blood urea nitrogen value. A third of studies (30 [33%] studies) measured kidney function by more than one method.
The overall prevalence of CKD from the 21 medium-quality and high-quality studies was 13·9% (95% CI 12·2–15·7; ? gure 2). Estimates ranged from 2% in Cote d’Ivoire to 30% in Zimbabwe. Nearly half (n=6) the countries represented had prevalence estimates ranging between 4% and 16%, and all countries but one (Zimbabwe) had prevalence estimates of less than 20% (? gure 2). In the case of Zimbabwe, the estimate is derived from only one study that included many people with HIV who had not received antiretroviral treatment, which probably explains the unusually high prevalence.29 The only other study to have measured a prevalence of CKD in Zimbabwe was a convenience sample of people with diabetes and was not included in the meta-analysis due to its low quality. We formally noted heterogeneity across the medium-quality and high-quality studies (n=21; I 2=47·11%, p=0·0009; ? gure 2). Heterogeneity across a ll s tudies w as m uch h igher (I 2=67·64%, p =0·0051).or many countries (Malawi, Burundi, Ethiopia, Tanzania, Benin, Burkina-Faso, Guinea, and Senegal) we detected no medium-quality or high-quality estimates for the prevalence of CKD. F or instance, although nine studies were done in Ethiopia, including more than 4000 people, all were convenience samples in people with diabetes or HIV (table). Additionally, even though we report estimates for Sudan, Nigeria, and South Africa, most studies from these countries were not included in the combined analysis due to issues with convenience sampling of high-risk populations and inconsistent measurements of kidney function. Of the three studies that met criteria for the highest quality grade, one was from Ghana and two were from the Democratic Republic of Congo (DRC). Both studies from DRC were done in an urban setting between 2006 and 2007 and reported proteinuria prevalence to be 5% and 17%, respectively.45,46 Only one of the studies estimated
GF R, and it showed that the prevalence of CKD when
de? ned as a GFR of less than 60 mL per min per 1·73 m2
by the MDRD formula was 7·8% (95% CI
7·6–8·1).45 The one high-quality study from Ghana was done in a rural population in 2010 and reports a CKD prevalence of 30% when de? ned as a GF R of less than 90 mL per min per 1·73 m2 by the MDRD formula but only 1·6% when de? ned as a GF R of less than 60 mL per min per 1·73 m2.47 The method of measuring kidney disease also a? ected the prevalence estimates across countries (table). Measurement of urine protein was the most common
method of determining the presence of kidney disease
(69%). Among medium-quality and high-quality studies, the prevalence of proteinuria ranged between 6% in
Sudan to 24% in Zambia. The estimate from Cameroon is derived by pooling two arms (one urban and one rural) from one case-control study that was designed to
compare the prevalence of nephropathy in an onchocerciasis-endemic area, and the estimate from
Ghana is derived from only one medium-quality study done in an urban population that was speci?
cally designed to measure end-organ damage such as proteinuria in people with hypertention.41,43 The one high-quality study from Ghana did not measure proteinuria.47Among the studies reporting both Cockcroft-Gault and MDRD formulae (n=10), estimates derived from the Cockcroft-Gault formula were higher in every study, although not signi? cantly so (p=0·080), with the di? erence less apparent for people with a creatinine clearance and GFR between 60 and 90 (p=0·170). Only two studies, one medium-quality study from Rwanda and one high-quality study from Ghana, reported estimated GFR by the CKD-EPI formula, and they reported estimates that were similar to those reported by the MDRD formula (p>0·900 for
di?
erence) at a GFR of less than 60 mL per min per 1·73 m2 and GFR 60–90 mL per min per 1·73 m2.27,47 We could not further characterise the extent to which people with an estimated GFR of less than 60 mL per min per 1·73 m2 had concurrent proteinuria. The table shows the results for each country strati? ed by comorbidity. Among people with HIV, the prevalence of CKD was highest in Nigeria, Zimbabwe, and Uganda (table 1). The remaining countries of Cameroon, C?te d’Ivoire, Kenya, Mozambique, Rwanda, South Africa, and Zambia had prevalence estimates of 1–10%. The di? erence between these two sets of countries might be related to the treatment status of the HIV participants. Only 16% of the participants in the studies from Nigeria, Zimbabwe, and Uganda were receiving care for their HIV, whereas all participants from the remaining countries were receiving some form of care for HIV.In people with diabetes or hypertension in DRC, Ghana, Nigeria, Rwanda, South Africa, Sudan, and
Figure 2: Forest plot of random-e? ects meta-analyses showing the pooled prevalence of chronic kidney disease by country
Articles
Zambia, the prevalence of CKD ranged from 4% to 24% (table), with proteinuria almost exclusively used (96%) as the marker.
The vast majority of studies were done in an urban setting (83 [93%] studies). Of the seven studies done in rural settings, most were of medium quality (n=6)27,31,32,37,38,43 and one was high quality.47 Of the studies done in urban settings, only 13 (16%)met criteria for either medium28–30,33–36,39–42 or high quality.45,46 We detected no di?erence in the prevalence of CKD between urban (12·4%, 95% CI 11–14) and rural (16·5%, 13·8–19·6) settings (p=0·474).
Examination of the prevalence by decade suggested that CKD prevalence increased in each of the four previous decades. However, de?nitive conclusions are limited by low study quality, detection bias, and publication bias that prevent meaningful analysis. Only eight studies were done before 1980 and only 31 studies were done between 1980 and 1999; 57 studies were done in the 2000s. Before the year 2000, only ? ve medium-quality studies and no high-quality studies were done.28,35,39,40,43
Discussion
Our systematic analysis of available studies showed that CKD is a prevalent and potentially escalating disease across sub-Saharan Africa with risk factors that include both communicable and non-communicable diseases. Although use of di? erent measurement and sampling methods limit direct comparison between studies,
Articles
examination of the medium-quality or high-quality studies suggests that the estimated prevalence in many sub-Saharan African countries might approximate those in other low-to-middle-income countries in southeast Asia such as Taiwan, Vietnam, Malaysia, and India as well as the Latin American countries of Mexico, Brazil, and Uruguay.48–53 Additionally, it also implies that CKD might not be a burden exclusive to high-income regions.54 Poor quality data for the prevalence of CKD in sub-Saharan Africa was pervasive, and our review identi? ed that the major reasons were related to poor sampling methods, unreliable kidney function measurements, and variable data availability in each country. Together, our ?ndings draw attention to an important limitation to systematically addressing an increasing health problem throughout sub-Saharan Africa.
Non-communicable diseases are thought to be a rising global epidemic that disproportionally a?ect the economic, social, and health outcomes of resource-poor countries and low-income populations.55 F or sub-Saharan Africa, conservative projections for the year 2030 predict that 18·65 million people will have diabetes, and predictions for hypertension, obesity, and tobacco use are similarly alarming.56–59 In the context of CKD, these are portentous forecasts in view of the high burden of CKD in people with diabetes or hypertension shown in our analysis. Communicable diseases such as HIV might also be contributing to the large burden of CKD in sub-Saharan Africa. Management of chronic diseases in people living with and receiving treatment for HIV is becoming a challenge because fewer people die from HIV-related complications and more develop chronic diseases. This recognition has led to a strong international commitment to integrate HIV with the care of non-communicable diseases.60 Our ?ndings suggest that CKD is among the important potential chronic complications of HIV in sub-Saharan Africa. HIV treatment for the 22 million people with the infection in sub-Saharan Africa has become an urgent issue within the global community. However, our ? ndings suggest that the prevalence of CKD is substantial even in people receiving care for HIV, which further substantiates the call to integrate HIV treatment with the care of non-communicable diseases.14
An important di?culty in the detection and management of CKD in sub-Saharan Africa is the absence of reliable and validated measures of kidney function. Our ?ndings show that, depending on the method of measurement or de?nition of kidney impairment, prevalence estimates can vary substantially even within similar populations, and although the CKD-EPI and modi?ed CKD-EPI formulae have been suggested to most closely approximate GF R in many African populations, they have been applied in only two prevalence studies and remain unvalidated.27,47,61 Proteinuria is another important and easily obtained clinical indicator of glomerular disease and might portend increased cardiovascular risk in people with HIV, diabetes, or hypertension.However, the best method of urinary protein detection remains unknown, and our ? ndings showed that the variability and dearth of available data are major limitations. Nonetheless, our results still suggest that, when measured, the prevalence of proteinuria in people with diabetes, hypertension, or HIV is substantial, which again stresses a need for the integration of HIV and non-communicable diseases.
The major strength of our study was the comprehensive inclusion of many studies with a transparent assessment of CKD prevalence in sub-Saharan Africa. Because of the shortage of high-quality data, we chose not to limit our systematic review to any particular population, time period, or measurement of kidney disease. We saw much heterogeneity between studies, which is a common ? nding in many meta-analyses. Wherever possible, we presented the prevalence estimates descriptively in many di? erent contexts to account for this heterogeneity, and we believe that exclusion of the low-quality studies from the review would have provided a restricted and false representation of the full state of renal epidemiology in sub-Saharan Africa. However, because such heterogeneity presented a less than ideal setting for pooling of data, we limited the meta-analyses to only studies of medium or high quality, and we felt that the absence of continent-wide epidemiological data for kidney disease justi? ed the method. The lack of reliable measures of kidney function was a major limitation of systematically reviewing the prevalence of kidney disease in sub-Saharan Africa, but this ?nding has drawn attention to the large gap in reliable and validated estimates of kidney function. Additionally, we did not review articles in non-indexed journals, which might have introduced a slight publication bias.
There are many other implicated causes of CKD in sub-Saharan Africa that we were not able to analyse, including chronic glomerulonephritis and the widespread use of herbal or traditional medication. Up to 80% of the populations in sub-Saharan countries are estimated to use herbal or traditional medicines, which are thought to have been associated with 35% of all new cases of acute kidney injury. Some of the most commonly used toxic herbs used as drugs are Securidaca longepedunculata (African violet tree), Cape aloes,and Eup h orbia matabalensis,but few follow-up studies have been done on kidney disease caused by these agents.62–64
On a continent where chronic diseases and cardiovascular deaths are now drawing increasing attention, CKD should not be overlooked. Approaches to addressing the rising epidemic of non-communicable diseases might initially include screening for CKD in speci? c high-risk populations, especially in view of the fact that cardiovascular disease and chronic kidney disease share many common risk factors.58,65,66 Such might be especially important when considering that the incidence of end-stage renal disease, an almost uniformly
Articles
fatal disease in sub-Saharan Africa, mirrors the prevalence of non-communicable diseases in high-income countries.7,16 Although poor data quality limit de? nitive inferences, our ? ndings suggest that CKD is a substantial health burden in sub-Saharan Africa with many risk factors, including many communicable and non-communicable diseases. These ?ndings are of global interest and should signal a call to action. Contributors
JWS contributed to protocol design, study design, the literature
review, quality assessment, data extraction, statistical analysis, data interpretation, article preparation, article review, and correspondence.
BJ contributed to statistical analysis, article preparation, and article review. ST contributed to the literature search, quality assessment,
data extraction, data interpretation, article preparation, and article review. NH contributed to the literature search, language translation, quality assessment, data extraction, and article review. RM contributed
to the literature search and data extraction. SN contributed to article preparation and review. UP contributed to protocol design, study design, data interpretation, and article preparation and review.
Con? icts of interest
We declare that we have no con? icts of interest.
Acknowledgments
We thank Megan von Isenburg for her help in generating our database search.
References
1 United Nations General Assembly. Political declaration of the
high-level meeting of the general assembly on the prevention and
control of non-communicable diseases. https://www.wendangku.net/doc/8716253427.html,/en/ga/
president/65/issues/ncdiseases.shtml (accessed Feb 13, 2013).
2 WHO. Preventing chronic diseases: a vital investment. http://www.
who.int/chp/chronic_disease_report/en/ (accessed Jan 20, 2013).
3 Codreanu I, Perico N, Sharma SK, Schieppati A, Remuzzi G.
Prevention programmes of progressive renal disease in developing
nations. Nephrology 2006; 11: 321–28.
4 Lozano R, Naghavi M, Foreman K, et al. Global and regional
mortality from 235 causes of death for 20 age groups in 1990 and
2010: a systematic analysis for the Global Burden of Disease Study
2010. Lancet 2012; 380: 2095–128.
5 The World Bank. World Development Indicators.
https://www.wendangku.net/doc/8716253427.html,/region/SSA (accessed Oct 22, 2013).
6 Naicker S. Burden of end-stage renal disease in sub-Saharan Africa.
Ethn Dis 2009; 19: 1–13.
7 Idem. Epidemiology of ESRD: a world-wide perspective.
In: El Nahas M, ed. Kidney diseases in the developing world and
ethnic minorities. London, UK: Taylor & Francis, 2005: 1–13.
8 Population Division of the Department of Economic and Social
A? airs of the United Nations Secretariat. World population prospects: the 2007 revision and world urbanization prospects. http://www.
https://www.wendangku.net/doc/8716253427.html,/esa/population/unpop.htm (accessed Feb 3, 2013).
9 Thomas BA, Wulf S, Mehrotra R, Himmelfarb J, Naghavi M,
Murray CJ. The rapidly growing global burden of end-stage renal
disease—an analysis of the chance in maintenance dialysis
prevalence between 1990 and 2010. Atlanta: Proceedings of the
American Society of Nephrology, 2013.
10 Grassman AS, Gioberge S, Moeller S, Brown G. ESRD patients
in 2004: global overview of patient numbers, treatment modalities
and associated trends. Nephrol Dial Transplant 2005; 20: 2587–93.
11 Wyatt C, Klotman P, D’Agati V. HIV-associated nephropathy:
clinical presentation, pathology, and epidemiology in the era
of anti-retroviral therapy. Semin Nephrol 2008; 28: 513–22.
12 US Renal Data System. USRDS 2007 Annual Data Report: Atlas of
CKD and ESRD in the US. https://www.wendangku.net/doc/8716253427.html,/adr.aspx(accessed Dec 12, 2012).
13 Wyatt, C, Meliambro K, Klotman P. Recent Pprogress in
HIV-associated nephropathy. Annu Rev Med 2012; 63: 1–13.
14 Joint United Nations Programme on HIV/AIDS. Report on the
global HIV/AIDS epidemic. https://www.wendangku.net/doc/8716253427.html,/en/
KnowledgeCentre/HIVData (accessed Dec 12, 2012).15 Katz I, Gernholtz T, Naicker S. Africa and nephrology: the forgotten
continent. Nephrol Clin Pract 2011; 117: 320–27.
16 Luyckx VA, Naicker S, McKee M. Equity and economics of kidney
disease in sub-Saharan Africa. Lancet 2013; 382: 103–04.
17 Liberati A, Altman DG, Tetzla? J, et al. The PRISMA statement
for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration.
PLoS One 2009; 6: e1000100.
18 Madhukar P, McCulloch M, Gormon J, et al. Systematic reviews
and meta-analyses: an illustrated, step-by-step guide.
National Med J India 2004; 17: 86–95.
19 Guyatt GH, Rennie D. Users’ guides to the medical literature:
a manual for evidence-based clinical practice.Chicago: AMA Press,
2002.
20 Juni P, Altman DG, Egger M. Systematic reviews in health care:
assessing the quality of controlled clinical trials. BMJ 2001; 323: 42–46.
21 Whiting P, Rutjes A W, Reitsma JB, Bossuyt PM, Kleijnen J.
The development of QUADAS: a tool for the quality assessment
of studies of diagnostic accuracy included in systematic reviews.
BMC Med Res Methodol 2003; 3: 25.
22 Shamliyan T, Kane R, Dickinson S. A systematic review of tools
used to assess the quality of observational studies that examine
incidence or prevalence and risk factors for disease.
J Clin Epidemiology 2010; 63: 1061–70.
23 Shamliyan, T, Kane RL, Ansari MT, et al. Development of quality
criteria to evaluate non-therapeutic studies of incidence,
prevalence, or risk factors of chronic diseases: pilot study of new checklist.Rockville, MD: Agency for HealthCare Research and
Quality, 2010.
24 DerSimonian R, Laird N. Meta-analysis in clinical trials.
Control Clin Trials 1986; 7: 177–88.
25 Wilson EB. Probable inference, the law of succession, and statistical
inference. J Am Stat Assoc 1927; 22: 209–12.
26 Newcombe R. Two-sided con? dence intervals for the single
proportion: comparison of seven methods. Stat Med 1998; 17: 857–72.
27 Wyatt CM, Shi Q, Novak JE, et al. Prevalence of kidney disease in
HIV-infected and uninfected Rwandan women. PLoS ONE 2011;
6: 1–6.
28 Rolfe M. Diabetic renal disease in central Africa. Diabet Med 1988;
5: 630–33.
29 St?hr W, Walker AS, Munderi P, et al. Estimating glomerular
? ltration rate in HIV-infected adults in Africa: comparison of
Cockcroft-Gault and Modi? cation of Diet in Renal Disease
Formulae. Antivir Ther 2008; 13: 761–70.
30 Abu-Aisha H, Elhassan EAM, Khamis AH, Abu-Elmaali A.
Chronic kidney disease in police forces households in Khartoum, Sudan: pilot report. Arab J Nephrol Transplant 2009; 2: 21–26.
31 Makulo R, Nseka MN, Jadoul M, et al. Albuminurie pathologique
lors du dèpistage du diabète en milieu semi-rural: citè de: Kisantu en RD Congo. Nephrol Ther 2010; 6: 513–19.
32 Arogundade F, Sanusi A, Hassan M, et al. Undiagnosed
hypertension and proteinuria in a market population in Ile-Ife,
Nigeria. Arab J Nephrol Transplant 2011; 4: 141–46.
33 Ayodele OE, Okunola OO, Afolabi MO, Oluyombo R,
Gbadegesin BA, Oyeleye AE. Prevalence of hypertension, diabetes, and chronic kidney disease in participants in the 2009 world kidney day screening exercise in southwest Nigeria. Hong Kong J Nephrol
2011; 13: 55–63.
34 Onyemelukwe GC, Rao R. Renal function in non-insulin dependent
diabetes mellitus. Cent Afr J Med 1991; 37: 150–52.
35 Akanji AO, Bella AF, Osotimehin BO. Cheiroarthropathy and long-
term diabetic complications in Nigerians. Ann Rheum Dis 1990;
49: 28–30.
36 Emem CP, Arogundade F, Sanusi A, Adelusola K, Wokoma F,
Akinsola A. Renal disease in HIV-seropositive patients in Nigeria: an assessment of prevalence, clinical features and risk factors.
Nephrol Dial Transplant 2008; 23: 741–46.
37 Rayner B, Becker P. The prevalence of microalbuminuria and ECG
left ventricular hypertrophy in hypertensive patients in private
practices in South Africa. Cardiovasc J S Afr 2006; 17: 245–49.
38 Jao J, Lo W, Toro P, et al. Factors associated with decreased kidney
function in HIV-infected adults enrolled in the MTCT-plus initiative in sub-Saharan Africa. J Acquir Immune De? c Syndr 2011; 57: 40–45.
Articles
39 Braude M, Osler C, Taylor D, Kalk WJ. Hypertension, proteinuria
and azotaemia in diabetes. S Afr Med J 1991; 2: 431–33.
40 Jackson WP, Goldberg MD, Major V, Campbell GD. Vascular and
other diabetes-related disorders among Natal Indian diabetics and non-diabetics. S Afr Med J 1970; 44: 279–85.
41 Addo J, Smeeth L, Leon DA. Hypertensive target organ damage in
Ghanaian civil servants with hypertension. PLoS One 2009; 4: 1–7. 42 Mulenga LB, Kruse G, Lakhi S, et al. Baseline renal insu? ciency
and risk of death among HIV-infected adults on anti-retroviral
therapy in Lusaka, Zambia. AIDS 2008; 22: 1821–27.
43 Ngu JL, Chatelanat F, Leke R, Ndumbe P, Youmbissi J. Nephropathy
in Cameroon: evidence for ? larial derived immune-complex
pathogenesis in some cases. Clin Nephrol 24: 128–34.
44 Afolabi MO, Abioye-Kuteyi EA, Arogandade FA, Bello IS.
Prevalence of chronic kidney disease in a Nigerian family practice population. S Afr Fam Prac 2009; 51: 132–37.
45 Sumaili EK, Krzesinski JM, Zinga CV, et al. Prevalence of chronic
kidney disease in Kinshasa: results of a pilot study from the
Democratic Republic of Congo. Nephol Dial Transplant 2009;
24: 117–22.
46 Sumaili EK, Nseka NM, Lepira FB, et al. Screening for proteinuria
and chronic kidney disease risk factors in Kinshasa: a world kidney day 2007 study. Nephron Clin Pract 2008; 111: 220–28.
47 Eastwood JB, Kerry SM, Plange-Rhule J, et al. Assessment of GFR
by four methods in Aadults in Ashanti, Ghana: the need for an
eGFR equation for lean African populations. Nephrol Dial Transplant 2010; 25: 2178–87.
48 Rajapurkar MM, John GT, Kirpalani AL, et al. What do we know
about chronic kidney disease in India: ? rst report of the Indian
CKD registry. BMC Nephrology 2012; 13: 10–18.
49 Vivekanand J. Human resources, training, and the growing
worldwide epidemic of kidney disease: current status of chronic
kidney disease care in southeast Asia. Semin Nephrol 2009;
29: 487–96.
50 Hwang SJ, Tsai JC, Chen HC. Epidemiology, impact, and
preventative care of chronic kidney disease in Taiwan.
Nephrology (Carlton) 2010; 15 (suppl 2): 3–9.
51 Amato D, Alvarez-Aguilar C, Castaneda-Limones R, et al.
Prevalence of chronic kidney disease in an urban Mexican
population. Kidney Int 2005; 97 (suppl): 11–17.
52 Schwedt E, Sola L, Rios PG, Mazzuchi N. Improving the
management of chronic kidney disease in Uruguay: a national renal healthcare program. Nephron Clin Pract 2010; 114: 47–59.53 De Lima AO, Kesrouani S, Gomes RA, Cruz J,
Mastroianni-Kirsztain G. Population screening for chronic kidney disease: a survey involving 38 721 Brazilians. Nephrol Dial Transplant 2012; 27 (suppl): 135–38.
54 Coresh J, Selvin E, Stevens LA, et al. Prevalence of chronic kidney
disease in the United States. JAMA 2007; 298: 2038–47.
55 United Nations O? ce of the Secretary General. Report on the
prevention and control of non-communicable diseases: follow-up of 19th UNAIDS Conference. https://www.wendangku.net/doc/8716253427.html,/en/ga/president/65/ issues/ncdiseases.shtml (accessed Feb 13, 2013).
56 Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of
diabetes. Diabetes Care 2004;27: 1047–53.
57 Kearny PM, Whelton M, Reynolds K, et al. Global burden of
hypertension: analysis of worldwide data. Lancet 2005; 365: 217–23.
58 Haslam DW, James WP. Obesity. Lancet 2005;366: 1197–209.
59 Wang Y, Chen X, Song Y, Caballero B, Cheskin LJ. Association
between obesity and kidney disease: a systematic review and
meta-analysis. Kidney Int 2008;73: 19–33.
60 The International AIDS Conference. Report of the 19th Annual
AIDS Conference. Washington DC: IAC, 2012.
61 Earley A, Miskulin D, Lamb EJ, Levey AS, Uhlig K. Estimating
equations for glomerular ? ltration rate in the era of creatinine
standardization: a systematic review. Ann Intern Med 2012; 5: 785–95.
62 Colson C, DeBroe ME. Kidney injury from alternative medicines.
Adv Chron Kidney Dis 2005; 12: 261–75.
63 Bagnis CI, Deray G, Baumelou A, et al. Herbs and the Kidney.
Am J Kidney Dis 2004; 44: 1–11.
64 Adelkum TA, Ekwere TR, Akinsola, A. The pattern of acute toxic
nephropathy in Ife, Nigeria. West Afr J Med 1999; 18: 60–63.
65 Coresh J, Wei GL, McQuillan G, et al. Prevalence of high blood
pressure and elevated serum creatinine level in the United States: Findings from the third National Health and Nutrition Examination Survey (1988–94). Arch Intern Med 2001; 161: 1207–16.
66 Levey AS, Beto JA, Coronado BE, et al. Controlling the epidemic of
cardiovascular disease in chronic renal disease: What do we know?
What do we need to learn? Where do we go from here?
Am J Kidney Dis 1998; 32: 853–905.