Ethnic minority disparities in progression and mortality of pre-dialysis chronic kidney disease: a systematic scoping review

Background There are a growing number of studies on ethnic differences in progression and mortality for pre-dialysis chronic kidney disease (CKD), but this literature has yet to be synthesised, particularly for studies on mortality. Methods This scoping review synthesized existing literature on ethnic differences in progression and mortality for adults with pre-dialysis CKD, explored factors contributing to these differences, and identified gaps in the literature. A comprehensive search strategy using search terms for ethnicity and CKD was taken to identify potentially relevant studies. Nine databases were searched from 1992 to June 2017, with an updated search in February 2020. Results 8059 articles were identified and screened. Fifty-five studies (2 systematic review, 7 non-systematic reviews, and 46 individual studies) were included in this review. Most were US studies and compared African-American/Afro-Caribbean and Caucasian populations, and fewer studies assessed outcomes for Hispanics and Asians. Most studies reported higher risk of CKD progression in Afro-Caribbean/African-Americans, Hispanics, and Asians, lower risk of mortality for Asians, and mixed findings on risk of mortality for Afro-Caribbean/African-Americans and Hispanics, compared to Caucasians. Biological factors such as hypertension, diabetes, and cardiovascular disease contributed to increased risk of progression for ethnic minorities but did not increase risk of mortality in these groups. Conclusions Higher rates of renal replacement therapy among ethnic minorities may be partly due to increased risk of progression and reduced mortality in these groups. The review identifies gaps in the literature and highlights a need for a more structured approach by researchers that would allow higher confidence in single studies and better harmonization of data across studies to advance our understanding of CKD progression and mortality.

attributed to faster progression of CKD and better CKD survival [7,8]. Understanding which ethnic groups have worse outcomes and which factors influence adverse outcomes can help clinicians and policy makers target health care efforts and resources and improve outcomes for individuals, as well as inform policies to reduce health inequities.
To our knowledge, no study has systematically scoped studies exploring the range of risk factors for ethnic differences in CKD progression and mortality. A systematic review published in 2010 [9] that investigated ethnic differences in CKD progression, and had similar inclusion criteria to the current study, identified 5 relevant studies and concluded little evidence for ethnic differences in CKD progression and a lack of appropriately designed studies to assess ethnic differences in CKD progression. However, there have since been further studies in the field. Furthermore, no studies have systematically synthesised evidence for ethnic differences in mortality for people with pre-dialysis CKD, which would further contribute to understanding of ethnic differences in CKD progressiona competing risk. A scoping review rapidly examines the extent, range, and nature of existing knowledge in a diverse body of literature, summarises research findings and identifies research gaps [10]. In contrast to systematic reviews, scoping reviews identify and synthesise the breadth of knowledge in a given area without in depth assessment of study quality. The aim of this scoping review was to identify and present findings from studies addressing ethnic differences in pre-dialysis CKD progression and mortality, the key factors that underpin these ethnic differences, and identify areas where further research is needed. Understanding the patterns and potential mechanisms in ethnic minorities in high-income countries may contribute to CKD prevention and care in countries where they are the main ethnic groups. We conducted a scoping review to synthesize the existing literature comparing CKD progression and mortality for ethnic minority and non-ethnic minority adults with pre-dialysis CKD.

Methods
Our scoping review was conducted in line with the fivestage framework outlined by Arskey and O'Malley (2005) [10]. This framework includes formulating the research question, identifying relevant studies, study selection, charting the data, and collating, summarizing and reporting results. We acknowledge that classification of ethnicity can be complex and challenging. In this study, ethnic minority was defined as belonging to a particular group of people with common social, physical, national, linguistic, cultural, ancestral backgrounds and other such attributes living in a country where they differ from the majority [11]. Ethnic minority groups include African-Americans, Hispanic, Asian (East Asia, Southeast Asia, Indian subcontinent), Native Hawaiian or Pacific Islanders, and American Indian or Alaska Native in US studies, and Afro-Caribbean, South Asian (comprising Indians, Pakistanis, Bangladeshis), East Asians [e.g.: Chinese], and other Asian countries, and First Nation populations in UK and other countries [e.g.: Aboriginal Australians]. Pre-dialysis CKD was defined by glomerular filtration rate (GFR) or a combination of urinary albumin to creatinine ratio and GFR, and not requiring RRT.

Information sources and search strategies
Search terms for ethnic groups included 'ethnic groups, race, minority, Asian, Caucasian, Hispanic, Continental population groups, and African' [11]. Search terms for CKD included 'kidney diseases, renal insufficiency, glomerular filtration rate'. Searches were expanded using truncation symbols and search terms were combined using Boolean operators. The electronic databases Medline OVID, Embase, CINAHL, PsycINFO, Web of Science, Scopus, Social Care Online, Applied Social Sciences Index and Abstracts (ASSIA), and Cochrane Database of Promoting Health Effectiveness Reviews were searched from 1992 to July 2017. An updated search was conducted in February 2020 to identify more recent eligible studies. A time frame of 1992 onwards was set to capture evidence from the last 28 years and the searches were limited to the English language. Searches were conducted without a study design filter. Bibliography searches of key papers were performed. The search strategies for each database can be found in Appendix 1.

Article selection
A comprehensive and iterative approach to the literature searches for evidence was taken to ensure that a broad range of perspectives was captured. Articles were included in the review if the following criteria were met: (1) used an adult study population and (2) compared risk of progression [e.g., (decline in) (estimated) glomerular filtration rate (GFR)] and/or mortality for ≥2 ethnic groups with pre-dialysis CKD. Articles that did not compare outcomes between ethnic minority and non-ethnic minority groups or focused on dialysis/transplant patients were excluded. Two reviewers (HH and RH) reviewed articles to determine eligibility for inclusion. Any discrepancies were resolved by discussion with a third reviewer (SF).

Data extraction and synthesis
Data were extracted into a series of evidence tables, developed a priori, by one reviewer (HH). One evidence table was produced for each outcome group. Each table included details on the authors, date of publication, country in which the study was conducted, study aims, ethnic groups included in the study, study design, outcomes of interest, and key findings on factors associated with any ethnic differences in CKD outcomes for each study. Evidence tables can be found in Appendix 2. A narrative synthesis approach was taken to summarise the evidence.

Search strategy, study selection and data extraction
The results of the search strategy and selection process are shown in Fig. 1. 8059 citations were identified from the search. After removing duplicates and title and abstract screening 227 studies met the criteria for full text review, from which 50 were selected for inclusion in the review. Our updated searches identified 5 relevant studies published between July 2017 and February 2020, resulting in a total of 55 studies included in the review. These 55 studies included 1 systematic review [9] and 8 literature reviews on CKD progression [12][13][14][15][16][17][18][19]. Together these reviews included 19 of the individual studies identified in our searches. The reviews were published between 2004 and 2018, addressed slightly different research questions, and included different studies (Table 1). Only two of the reviews [9,19]

Indigenous populations
Two studies, 1 conducted in Australia and 1 in Canada, focused on CKD progression in Indigenous populations [17,53]. One was a discussion paper and the other a prospective cohort study.

East Asians
Five cohort studies -3 conducted in the US and 2 in Canada-explored mortality differences for East Asians and Caucasians. All 5 studies reported lower risk of mortality for East Asians compared to Caucasians (adjusted HR (95% CI) ranging from 0.58 (0.52-0.65) to 0.69 (0.55-0.88)) [25,26,49,56,65]. Age and sex, proteinuria, blood pressure, diabetes, cardiovascular disease and medications (antihypertensive drugs and statins), and C-reactive protein [49,59] partly explained the lower risk of mortality observed for East Asians.

Pacific islanders
Three US studies, 1 prospective and 2 retrospective cohort studies, explored mortality differences for Asians/ Pacific Islanders and Caucasians [25,26,56]. Two reported lower risk of mortality for Pacific Islanders (adjusted HR (95% CI) ranging from 0.58 (0.52-0.65) to 0.76 (0.61-0.95) and 1 reported no significant differences. Factors contributing to ethnic differences in mortality for Pacific Islanders compared to Caucasians were not fully explored.

Hispanics
Six US studies examined ethnic differences in mortality for Hispanics and Caucasians [25,26,38,52,56,61]. Studies were either prospective or retrospective cohort studies. Three studies reported lower risk of mortality for Hispanics compared to Caucasians (adjusted HR (95% CI) ranging from 0. 66  Lower risk of mortality observed for Hispanics was partly explained by differences in urine protein levels [38] (with Hispanics having significantly lower risk of mortality than Caucasians at higher levels of urine protein but no significant ethnic differences at lower levels of urine protein) and not explained by differences in hypertension, diabetes, and use of medication including insulin [52].

Native Americans
One US retrospective cohort study compared mortality differences for Native Americans and Caucasians and reported higher risk of mortality for Native Americans (adjusted HR (95% CI):1.41 (1.08-1.84)) [56]. Factors contributing to mortality differences for Native Americans compared to Caucasians were not clear.

Discussion
We used an established and systematic methodology and searched a range of databases to capture the full range of existing studies on ethnic differences in pre-dialysis CKD progression and mortality. This scoping review identified evidence for higher risk of CKD progression in Afro-Caribbean/African-Americans, Hispanics, Asians compared to Caucasians which was at least partly explained by biological factors (e.g.: blood pressure) and comorbidities (such as diabetes, and cardiovascular disease), and lower risk of mortality for South and East Asians and Pacific Islanders compared to Caucasians. Our scoping review also identified mixed findings on risk of mortality for African-Americans and Hispanics compared to Caucasians. Future studies need to explore this, as studies reporting significant findings did not differ in the range of adjusted confounders compared to studies that found significant differences. The role of medication in the association between ethnicity and progression and mortality is complex, as differences in medication may represent unmet need in certain ethnic

Gaps in the literature
A key gap in this literature is understanding why Asians (South, East, and Pacific Islanders) and Hispanics live longer, despite having higher prevalence of comorbidities such as diabetes, cardiovascular disease, and heart failure. Future research may explore potential missing factors that may explain why these groups experience increased risk CKD progression but live longer. The search also identified most research on ethnic differences in CKD progression and mortality has been conducted in US populations and there is less research in other countries with a significant proportion of ethnic minority immigrant populations (e. There may be heterogeneity in findings within ethnic groups, as shown in a recent UK study that found risk of CKD progression was higher in Bangladeshis compared to Indians [36]. Similarly, identified studies did not distinguish or adjust for generational status or indigenous vs immigrant populations, though some existing studies may not have been captured through our searches. Some studies did not adjust for important confounders of the association between ethnicity and CKD outcomes. Firstly, very few (n = 15) of the included studies adjusted for socioeconomic status. This is particularly important for US studies, where low socioeconomic status is closely linked to ethnicity and independently associated with health insurance and access to health care. Secondly, most studies on CKD progression did not account for competing risk of death, so differences in progression in these studies may have been at least partly due to differences in survival across ethnic groups. A limited number of studies (n = 18) assessed ethnic differences in both progression and mortality. These studies seemed to suggest Hispanics and East and South Asians experience increased risk of CKD progression as result of lower competing risk of death, and poorer evidence for significant differences in risk of mortality for African-Americans and Hispanics compared to Caucasians. However, further research is needed to confirm these findings. Thirdly, there were few studies exploring genetic risk factors for CKD progression and mortality. Some studies have suggested other genetic risk factors such as genes encoding non-muscle myosin heavy chain type II isoform A for ESKD in African-Americans [66], but there is a lack of genetic studies comparing risk of ESKD across different ethnic groups with pre-dialysis CKD. Future studies may also explore whether differences between ethnic groups hold across countries or whether they differ due to societal and health care reforms. Fourthly, some studies did not assess progression and mortality stratified by level of CKD severity, making it difficult to directly compare or identify if ethnic differences in progression and mortality vary at different stages of CKD. An updated systematic review and meta-analysis that goes beyond scoping to assess bias in these studies and pool together estimates from the different studies (where possible) may help explain some of the mixed findings and improve our understanding of the extent and key predictors of ethnic differences in CKD outcomes. Finally, there was lack of data on level of control of biologic factors such as blood pressure and glycemia, as well as limited data on medication and adherence, and how these vary across ethnic groups, all of which are important for differences in CKD progression and mortality. Future studies should aim to capture this data as much as possible.

Limitations
The review was based on a comprehensive search of the literature. However, it is possible that some relevant studies may have been missed as the search was restricted to studies that were published in English and published after 1992. Studies that examined CKD progression and outcomes for only one ethnic group and did not make comparisons with another ethnic group were also excluded. Furthermore, there was limited additional searching of grey literature, though we believe the majority of relevant studies will have been captured through the different databases and bibliography.
An important limitation of the scoping review approach is that papers are not critically appraised in detail and quality of the individual studies is therefore not assessed [10]. However, this scoping review was based on established methodology [10] and has mapped the existing literature on ethnic differences in CKD outcomes, identified gaps in the research and highlighted the need for further systematic reviews and additional primary research focusing on cardiovascular-related and other adverse outcomes for pre-dialysis CKD.

Clinical and policy implications of this scoping review
Increased risk of CKD progression in ethnic minority groups may be tackled through closer monitoring and management of renal comorbidities such as diabetes and cardiovascular disease, for example through proteinuria and blood pressure measurement, particularly in these high risk groups. There has been some evidence to suggest incentivisation programs, such as the Quality and Outcomes Framework programme in the UK, may help improve care for diabetics with CKD [67]. Interventions including the use of medications such as reninangiotensin-aldosterone system blockers and patientprovider education interventions may also reduce risk of progression in high-risk groups [68]. In the UK, a national quality improvement programme has mapped laboratory data taken from all settings to derive graphs of kidney function over time. Declining kidney function is then flagged by a laboratory scientist and sent to primary care doctor for clinical review and referral, where necessary [69]. However, a better understanding of risk factors for CKD progression in high risk groups is needed to help develop more effective and targeted interventions.

Conclusions
Scoping reviews are a relatively novel method of systematically assessing a wide range of literature in a particular field, in order to identify important gaps in the literature, and inform more targeted systematic reviews or further studies. This is the first synthesis of the extensive body of literatures on ethnic differences in CKD progression and mortality. The findings of this review suggest higher rates of RRT in ethnic minority groups may be partly due to increased risk of progression and reduced mortality in these groups (compared to Caucasians), though more evidence is needed for African-American and Hispanic ethnicity. The review highlights the need for further studies using similar approaches and adjusting for the same confounders, which would improve our understanding of disease progression and mortality in people with CKD.