Time Trends and Causes of Infection-Related Mortality Among Patients Starting Dialysis in Finland: A Nationwide Cohort Study.

RATIONALE &

OBJECTIVE

Previously, we reported a decrease in infection-related mortality in kidney transplant recipients. Regarding patients treated with dialysis, it is unclear whether infection-related mortality is decreasing. Therefore, we investigated current time trends and specific causes of infection-related mortality over 2 decades in a large cohort of patients treated with dialysis.

STUDY

DESIGN

A nationwide cohort study.

SETTING &

PARTICIPANTS

Patients starting kidney replacement therapy were identified through the Finnish Registry for Kidney Diseases. Follow-up continued until death of any cause, kidney transplantation, moving abroad, recovery of kidney function, loss of follow-up, or end of study.

EXPOSURE: Long-term kidney replacement therapy.

OUTCOME: Death due to infection.

ANALYTICAL APPROACH: Incidence rates, incidence rate ratios, and adjusted Cox regression hazard ratios for infection-related deaths were calculated by sub-cohorts consisting of patients whose kidney replacement therapy was started either 2000-2004, 2005-2009, 2010-2014, or 2015-2019. As sensitivity analyses, we studied infection-related mortality within 1 year of dialysis initiation and performed competing risk analysis.

RESULTS

A total of 9,671 adult patients started long-term dialysis from 2000 to 2019. Infection-related deaths declined from 47 to 23 deaths per 1,000 person-years over the four 5-year intervals from 2000-2004 to 2015-2019. The hazard ratio of a Cox model including identified risk factors was 0.49 (95% CI, 0.39-0.62) for patients who started dialysis in 2015-2019 compared with those who started in 2000-2004. The most common specific causes of infection-related deaths were septicemia (38%), pulmonary infection (36%), and peritonitis (8%), whereas opportunistic infections rarely caused death.

LIMITATIONS: Death certificates may have low sensitivity for infectious diseases. Only one cause of death is available. Categories of infectious deaths may overlap.

CONCLUSIONS

Risk of dying due to infections has halved since the beginning of the millennium despite aging among patients treated with dialysis. The reason for this development requires further studies.