The number of patients with end-stage kidney disease (ESKD) increased tremendously causing increased health burden worldwide [1,2]. The number of ESKD patients who begin dialysis therapy doubled in the last 10 years in Korea [3]. The increase in diabetic patients and the elderly is one of the many causes contributing to the increase in the dialysis population [4,5].

The ESKD patients not only cause increased medical expenses but also demonstrate a high morbidity and mortality rate. Although dialysis treatment improved the survival rate of ESKD patients in past decades [6,7], hemodialysis (HD) patients have a much higher mortality rate than the age- and sex-matched general population [8]. The mortality rate from ESKD is even higher than that of regional breast cancer, colon cancer, and kidney cancer [8].

Age-specific life expectancy is a commonly used measure to estimate the impact of health status and disease burden at a population level [9]. Policymakers can use it to identify disease conditions that have the greatest impact on overall life expectancy and to prioritize specific health programs or funding [10]. Although there are a few research papers investigating risk factors and mortality trends in Korean maintenance HD patients [3,11], there has been no study calculating the remaining lifespan among maintenance HD patients in Korea. Therefore, this study was performed to calculate the remaining life expectancy among Korean maintenance HD patients and compare it to those of the general population as well as HD patients from other countries.

Universally, the incidence and prevalence of ESKD patients on renal replacement therapy increased while the mortality rate decreased over decades [3,14,15]. The possible reasons for the longevity of HD patients not only include improvement in dialysis-related technologies and new drugs but also may include improvement in the dialysis environment and policy [2,16]. Nevertheless, ESKD patients still demonstrate a high rate of hospitalization and mortality due to decreased immune function and multiple comorbidities.

Our study showed that HD patients have a shorter life expectancy compared to the general population. A previous study by the Italian Dialysis and Transplantation Registry also demonstrated that the dialysis population has shorter relative survival compared to the general population [8]. Our study also demonstrated that male patients have shorter survival compared to females, which is in line with the previous studies performed in Japan [13,17]. The differences in life expectancy between sexes became negligible in HD patients older than 80 years. This finding is also in line with the previous Japanese study [13]. However, interestingly, American or European HD patients did not show differences in life expectancy between the sexes. Indeed, American female HD patients even showed shorter remaining life expectancys at younger ages (Table 3). The previous study performed among the United States incident HD patients using data from the Dialysis Outcomes and Practice Patterns Study also demonstrated no difference between sexes in mortality [18]. The different phenomena between Asian and Western countries may be due to ethnic or genetic differences. However, this cannot be elucidated further in this study.

Our study also demonstrated that the remaining life expectancy was much shorter in American or European countries compared to Asian countries. Compared to the Japanese study, Korean HD patients showed a similar survival profile [13]. This can be explained by ethnic or genetic differences. The previous studies performed in the United States and the United Kingdom also demonstrated that white HD patients showed inferior survival compared to non-whites [14,18–21]. In the general population, non-whites, especially African Americans show a poorer survival rate. However, in the dialysis population, paradoxical survival benefit was shown in African Americans compared to Caucasians. It may be due to the younger age to start dialysis in African Americans or higher body mass index or nutritional index in them [19].

Lastly, our study showed a shorter life expectancy in patients with diabetic nephropathy. This is in line with USRDS or ERA-EDTA (European Dialysis and Transplant Association) annual reports [14,15]. However, the differences in remaining life expectancy decreased as the patient’s age increased. This suggests that the cause of ESKD is less influential than the age factor.

Our study has some limitations. Firstly, we did not include HD patients younger than 30 years. They may have greater survival when included. Secondly, we could not compare the remaining life expectancy according to the cause of ESKD in the population with extreme age due to a small number of patients in those age groups. Thirdly, other variables such as smoking, alcohol intake, exercise, or dialysis vintage were not adjusted in the analysis of life expectancy. The variables such as smoking, alcohol intake, or exercise were not collected from the HD quality assessment. The dialysis vintage was not adjusted when calculating life expectancy because when we include the variable, the number of patients in each age group becomes too small for comparison. Since we only included prevalent HD patients and we know the mortality rate is the highest in the first 3 months of HD treatment [22,23], we may underestimate the actual mortality rate. In international comparison, data from other countries comprise both HD and peritoneal dialysis populations, but our data only includes the HD population. In addition, initial data collection includes outpatient HD data and excludes hospitalized or transferred cases. Data from Japanese, European, and American include both inpatient and outpatient data. Also, the data collection method and period were quite different by country. This may lead to selection bias since outpatients may represent a ‘healthier’ population compared to hospitalized patients. Therefore, better life expectancy in Korean HD patients can be the result of selecting a healthier total cohort. However, the Japanese study analyzed data from both inpatients and outpatients and still demonstrated better outcomes compared to Western countries. In addition, our follow-up survival data includes all possible data regardless of hospitalization. We assume that genetic or ethnic factors may weigh heavier than the selection bias. We could not validate our result using statistical methods except the lifetable method. Finally, we could not assess the improvement in survival rate among HD patients over the years. This should be further elucidated in the future study.

Despite many limitations, this is the first study to assess the remaining life expectancy in the prevalent HD patients in Korea. This study can help clinicians explain the prognosis to their HD patients. Longitudinal analysis should be warranted to analyze the effect of advanced dialysis technology on improved survival rates among the HD population.