Background

Renal failure represents a growing but mostly undocumented cause of premature mortality in low-income and middle-income countries. We investigated changes in adult renal failure mortality and its key risk factors in India using the nationally representative Million Death Study.

Methods

In this cross-sectional analysis of population-based data, two trained physicians independently assigned underlying causes to 150 018 deaths at ages 15–69 years from a nationally-representative mortality survey in India for 2001–03 and 2010–13, using the International Classification of Diseases, 10th version (ICD-10). We applied the age-specific proportion of renal failure deaths for the 2010–13 period to the 2015 UN estimates of total deaths in India and calculated age-standardised death rates for renal failure by rural or urban residence, state, and age group. We used proportional mortality of renal deaths (cases) to injuries (controls) to calculate the odds of renal death in the presence of different comorbidities and stratified risks by decade of birth.

Findings

In 2001–03, 2·1% of total deaths among 15–69 year olds were from renal failure (1266 [2·2%] of 58 871; unweighted). By 2010–13, the proportion of deaths from renal failure had risen to 2·9% (2943 [3·2%] of 91 147; unweighted) of total deaths and corresponding to 136 000 renal failure deaths (range 108 000–150 000) of 4 688 000 total deaths nationally in 2015. Age-standardised renal death rates were highest in the southern and eastern states, particularly among adults aged 45–69 years in 2010–13. Diabetes, hypertension, and cardiovascular disease were all significantly associated with increased renal failure deaths, with diabetes the strongest predictor—odds ratio (OR) vs control 9·2 (95% CI 6·7–12·7) in 2001–03, rising to 15·1 (12·6–18·1) in 2010–13. In the 2010–13 study population, the diabetes to non-diabetes OR was twice as large in adults born in the 1970s (25·5, 95% CI 17·6–37·1) as in those individuals born during or before the 1950s (11·7, 9·1–14·9).

Interpretation

Renal failure is a growing cause of premature death in India. Poorly treated diabetes is the most probable reason for this increase. Strategies aimed at diabetes prevention, and early detection and treatment are urgently needed in India, as well as greater access to renal replacement therapy.

Funding

US National Institutes of Health, International Development Research Centre, Centre for Global Health Research, University of Toronto.

In many high-income countries the prevalence of chronic kidney disease approaches 15% of the adult population, ¹ predominantly due to diabetic and hypertensive nephropathy, and poses a large medical and economic burden. ² Kidney disease and death from renal failure is also a major, but poorly documented burden for low-income and middle-income countries. In low-income countries, infectious causes, environmental pollutants, and toxins are thought to be the primary causes of both acute kidney injury and chronic kidney disease. ³ Middle-income countries, such as India, which are undergoing rapid economic and epidemiological transitions appear to be facing a double burden, with continued high prevalence of infectious causes of kidney disease as well as rising rates of hypertension and particularly of untreated diabetes. ^{3 , 4 and 5}

In India, nationally representative, population-based data for renal failure incidence, prevalence, and deaths are absent at both a national and a subnational level. However, several factors suggest morbidity and mortality from renal failure is likely to be a growing public health concern for India. The prevalence of diabetes is estimated at 9% among adults in urban India, and is greater among those who are overweight or wealthy. ⁶ Hypertension prevalence is also rising in both urban and rural India. ⁷ Because the development of chronic kidney disease is insidious, and aetiologies such as diabetes and hypertension are often underdiagnosed at the population level, presentation with kidney disease is typically late; ⁸ for example, in one study, ⁹ end-stage renal failure accounted for about half of all chronic kidney disease presentations in India.

In this Article, we aimed to estimate renal failure mortality in the nationally representative Million Death Study (MDS) conducted in India, comparing data for 2001–03 and 2010–13. We estimate the number of adults deaths from renal failure, the contribution of diabetes, hypertension and cardiovascular disease, and changes over time in each period.

Research in context

Evidence before this study

We systematically searched PubMed and Embase databases with no specified start date up to Aug 1, 2016 for English language sources. We used the following PubMed search terms and translated appropriately for Embase: (“kidney disease” [MeSH], OR “renal failure”[MeSH]) AND (“mortality estimates” [tw], OR “burden of disease” [tw], OR “disease burden” OR “household” [tw] AND “survey” [tw]). We further delineated by low-income or middle-income countries, and specifically screened for representative, population-based studies of mortality due to renal failure. We found no nationally representative, population-based studies of renal failure mortality for any middle-income country. Population-based studies at the state and community level of renal failure incidence and prevalence exist for India, but any estimations of national mortality drawn from these are likely to be weak because these are typically not representatively sampled, and because as our study findings show, subnational age-standardised rates of renal failure mortality vary widely within India.

Added value of this study

To our knowledge, our study results provide the first nationally representative population-based estimate of deaths from renal failure in any low-income and middle-income country, and show that renal failure is a growing public health concern for India. We establish that much of the high burden and rising death rates are due to diabetes. Our estimates also show very low coverage of renal replacement therapy (dialysis and transplantation).

Implications of all the available evidence

Renal failure is an important and growing cause of premature adult mortality in India, driven by the rising prevalence of poorly treated diabetes, especially in urban areas. Access to dialysis and transplantation is poor and not commensurate with the large and rising burden of renal disease. Greater priority should be given to addressing renal failure as well as access to preventive and treatment services, including for diabetes, in India and other comparable middle-income countries.

Details of the MDS, including study design, physician assignment of the underlying cause of death, and statistical methods have been published elsewhere. ^{10 , 11 and 12} Briefly, the study uses an enhanced type of verbal autopsy method (a structured survey administered to a household member or close relative of the deceased by a trained non-medical field worker to record the signs and symptoms that occurred before death—this information is used to assign the most probable cause of death ¹³ ). From 2001 onwards, the MDS has monitored annual deaths in 1·3 million representative households within the Registrar General of India (RGI)'s Sample Registration System (SRS). The SRS partitions India into 1 million small areas after each census, from which 6671 small areas from the 2001 census and 7597 small areas from the 2011 census are randomly selected for continuous monitoring of household births and deaths. One of 800 non-medical RGI surveyors visits each house every 6 months and for any household with a death, interviews a family member or associate of the deceased and completes a two-page form ¹⁴ with structured questions and a half-page local language narrative that probes the presence or absence of key symptoms before death. For all deaths in people over 15 years of age, the surveyors ask the living respondent about the deceased's use of smoked or chewed tobacco and alcohol, and whether a doctor had ever diagnosed heart disease, stroke, hypertension, or diabetes. The field records are converted to electronic records and emailed independently to two of 400 specially trained physicians able to read the local language. Physician coding follows guidelines for the major underlying causes of deaths, coded using the International Classification of Diseases and Related Health Problems, 10th version (ICD-10). ¹⁵ ICD-10 coding differences undergo anonymous reconciliation by each of the two physicians, and persisting differences are adjudicated by a third physician. Random independent resampling of about 3% of deaths has shown cause-of-death results to be consistent with the RGI fieldwork. ^{10 and 16}