Risk Factors for Mortality at the Initiation of Maintenance Hemodialysis
Maintenance hemodialysis (MHD) is a cornerstone of renal replacement therapy for patients with end-stage renal disease (ESRD). Despite advancements in dialysis technology, mortality rates among MHD patients remain alarmingly high. Understanding the risk factors associated with mortality at the initiation of dialysis is critical for refining clinical management, guiding targeted interventions, and improving patient outcomes. This study retrospectively analyzed patients initiating hemodialysis at the Center for Hemodialysis in West China Hospital, Sichuan University, China, to identify prognostic factors influencing survival.
Study Design and Patient Population
The study employed a single-center, retrospective cohort design, enrolling patients who began MHD between July 2013 and July 2018. Inclusion criteria required patients to (1) receive regular hemodialysis for ≥3 months, (2) follow a standard bicarbonate dialysate regimen (three sessions weekly, 4 hours per session), and (3) receive enoxaparin calcium (60–80 IU/kg) as anticoagulation. Exclusion criteria eliminated patients under 18 years old, those with concurrent peritoneal dialysis, incomplete medical records, or a history of myocardial infarction or heart valve disease. A total of 311 eligible patients were included, with follow-up continuing until April 2019, patient death, withdrawal from dialysis, or transfer to another facility.
Baseline Characteristics
The cohort comprised 57% males, with a mean age of 53 years. Diabetes mellitus (75%) and arteriovenous fistula use (77%) were prevalent. Pneumonia was present in 15% of patients at dialysis initiation. Laboratory parameters at baseline included a mean hemoglobin of 90.0 ± 21.0 g/L, serum urea of 24.4 ± 11.5 mmol/L, cholesterol of 3.9 ± 1.1 mmol/L, and albumin of 39.4 ± 6.1 g/L. The average dialysis duration was 43 months.
Mortality Outcomes
Among 33 deaths (11% of the cohort), cardiovascular diseases (CVD) dominated as the leading cause (55%, n=18), followed by pneumonia (15%, n=5). Multiple organ failure and gastrointestinal bleeding each accounted for 12% of deaths (n=4), while malignancy-related complications caused 6% (n=2). These findings underscore the critical role of CVD and infections in driving mortality among MHD patients.
Survival Analysis and Pneumonia Impact
Kaplan-Meier survival curves stratified by pneumonia status revealed stark differences in outcomes. Patients with pneumonia had significantly shorter survival times compared to those without (P<0.001, log-rank test). Although median survival was not reached in either group, the divergence in curves highlighted pneumonia as a potent predictor of mortality.
Univariate and Multivariate Cox Models
Univariate analysis identified several factors associated with mortality, including older age (HR 1.048 per year, P<0.001), diabetes (HR 3.167, P=0.001), pneumonia (HR 3.775, P<0.001), elevated cholesterol (HR 1.308 per mmol/L, P=0.022), and lower albumin (HR 0.916 per g/L, P<0.001). Longer dialysis duration (HR 0.846 per month, P<0.001) and arteriovenous fistula use (HR 0.254, P<0.001) were protective.
After adjusting for age, the associations of diabetes, urea, and cholesterol with mortality attenuated, suggesting confounding by age. In the fully adjusted multivariate Cox model, three factors retained significance:
- Age: Each additional year increased mortality hazard by 2.8% (aHR 1.028, 95% CI: 1.000–1.056).
- Pneumonia: Comorbid pneumonia tripled mortality risk (aHR 3.179, 95% CI: 1.454–6.953).
- Cholesterol: Every 1 mmol/L increase elevated hazard by 44% (aHR 1.440, 95% CI: 1.077–1.925).
Protective factors included longer dialysis duration (aHR 0.831 per month, 95% CI: 0.788–0.876) and higher albumin levels (aHR 0.937 per g/L, 95% CI: 0.889–0.987).
Hospitalization Risk Factors
Multivariate logistic regression identified pneumonia and age as key drivers of hospitalization. Pneumonia increased hospitalization odds over threefold (OR 3.128, 95% CI: 1.718–5.696), while each year of age raised odds by 4.1% (OR 1.041, 95% CI: 1.019–1.063). These findings emphasize the clinical and economic burden of infections and advanced age in MHD populations.
Clinical Implications and Pathophysiological Insights
Cardiovascular Disease and Aging
CVD accounted for over half of all deaths, aligning with global data showing CVD as the leading cause of mortality in ESRD. Age-related physiological decline, compounded by uremic cardiomyopathy and accelerated atherosclerosis in ESRD, likely explains the heightened vulnerability of older patients. Notably, even younger MHD patients face substantial CVD risks, warranting aggressive cardiovascular risk modification regardless of age.
Cholesterol and Mortality
The paradoxical association of higher cholesterol with increased mortality contrasts with general population studies, where hypercholesterolemia primarily drives atherosclerosis. In MHD patients, elevated cholesterol may reflect malnutrition-inflammation complex syndrome (MICS), a state linked to poor prognosis. Statin therapy, beneficial in early chronic kidney disease (CKD), shows equivocal effects in dialysis populations, highlighting the need for tailored lipid management strategies.
Albumin as a Marker of Nutrition and Inflammation
Hypoalbuminemia, a hallmark of protein-energy wasting and inflammation, independently predicted mortality. Each 1 g/L decrease in albumin elevated mortality risk by 6.3% (aHR 0.937), underscoring the importance of nutritional interventions and inflammation control. Protocols to optimize dietary protein intake, address uremic anorexia, and manage comorbid conditions (e.g., infections) may mitigate this risk.
Pneumonia: A Modifiable Risk Factor
Pneumonia contributed disproportionately to mortality and hospitalization, consistent with US Renal Data System (USRDS) reports. Immune dysfunction, fluid overload, and frequent healthcare exposures predispose MHD patients to severe infections. Preventive measures, including vaccination programs, meticulous volume management, and prompt antibiotic therapy, could reduce pneumonia incidence and its downstream consequences.
Study Limitations
The study’s retrospective design and single-center origin limit generalizability. The small sample size (N=311) and low event rate (33 deaths) restricted power to detect associations for less common risk factors. Unmeasured confounders, such as socioeconomic status, residual renal function, and medication adherence, were not assessed. Furthermore, the homogeneous Chinese population may limit applicability to other ethnic groups. Prospective, multi-center studies with larger cohorts are needed to validate these findings and explore additional predictors.
Conclusions
This study identifies age, pneumonia, and hypercholesterolemia as independent risk factors for mortality in MHD patients initiating dialysis. Conversely, longer dialysis vintage and higher albumin levels confer protection. Cardiovascular diseases and infections dominate as primary causes of death, emphasizing the need for integrated care addressing both non-infectious and infectious complications. Clinical strategies should prioritize pneumonia prevention, nutritional optimization, and aggressive management of dyslipidemia and CVD in this vulnerable population. Future research must address the study’s limitations through prospective designs and broader geographic representation to refine risk stratification and therapeutic approaches.
doi:10.1097/CM9.0000000000000719
Was this helpful?
0 / 0