Following expansion of Medicare coverage to patients with end-stage renal disease (ESRD) 40 years ago, the average number of hours per hemodialysis session has decreased from 6 hours in 1973 to 3.5 to 4 hours in 2010. The decline is explained in part by the improved efficiency of dialyzers; however, there are few data on the clinical consequences of shorter session duration on patients on maintenance hemodialysis.
Shailender Swaminathan, PhD, and colleagues recently conducted a retrospective cohort study utilizing data from dialysis centers in the United States that prescribe the same session duration to all patients initiating maintenance hemodialysis therapy, rather than selecting incident patients for varying session lengths. The researchers sought to compare mortality rates of patients treated in centers that used initial dialysis session durations of either ≥4 or 3 hours for all incident patients. Study results were reported in the American Journal of Kidney Diseases [2017;70(1):69-75].
Study participants were patients with ESRD initiating maintenance hemodialysis therapy from January 2006 to December 2010; follow-up continued through December 2012. The cohort included 39,172 patients from 852 facilities who initiated treatment for ≥4 hours and 47,721 patients from 631 facilities who initiated treatment for 3 hours. The primary outcomes of interest were 2- and 1-year mortality rates.
Patients treated at ≥4 hour facilities were approximately 3 years younger than those treated at 3 hour facilities. A larger proportion of patients treated at ≥4 hour facilities were black compared with patients treated at 3 hour facilities (39% vs 21%). The remainder of the individual-level characteristics were similar across the ≥4-hour and 3-hour facilities.
The primary facility-level differences between ≥4-hour and 3-hour centers were that the 3-hour facilities were eight percentage points less likely to accept transient patients, eight percentage points more likely to have evening dialysis sessions, 10 percentage points more likely to report isolation dialysis sessions, and 10 percentage points less likely to reuse dialyzers compared with ≥4-hour facilities. Zip codes for the locations of ≥4-hour facilities had a lower total number of patients compared with zip codes for locations of the 3-hour facilities (55 vs 71).
The number of deaths within 1 year of dialysis initiation was 5492 in the ≥4-hour group and 10,372 in the 3-hour group; at 2 years, the total number of deaths at the ≥4-hour group was 8945 compared with 15,624 deaths in the 3-hour group. The 1-year mortality rate for patients who initiated hemodialysis at ≥4-hour facilities was 15.2 per 100 person-years, compared with 25.0 per 100 person-years for those who initiated hemodialysis therapy at 3-hour facilities. The 2-year mortality rates were 13.1 per 100 person-years and 20.5 per 100-person-years, respectively.
In a proportional hazard model that adjusted only for the year of initiation, the 2-year hazard ratio [HR] of death among those treated at ≥4-hour facilities was 0.72 (95% confidence interval [CI], 0.66-0.78). Following adjustment for individual demographic and clinical characteristics, the HR was 0.78 (95% CI, 0.71-0.85). In a model that used inverse probability weights to reweight the data, the HR was 0.79 (95% CI, 0.73-0.86). HRs for 1-year mortality were 0.70 (95% CI, 0.64-0.76), 0.76 (95% CI, 0.69-0.83), and 0.77 (95% CI, 0.70-0.84), respectively.
Estimates of HRs and 95% CIs stratified by age, sex, race, and presence or absence of heart disease and diabetes were similar to those in the primary analysis, suggesting a consistent association of ≥4-hour initial session duration with lower morality compared with 3 hours of initial session duration.
Limitations cited by the authors included lack of observation of hemodialysis duration in sessions subsequent to initiation and only including patients treated in facilities with uniform session length at initiation for all their patients. Further, there was no available information on dialysis dosage and patients’ residual kidney function at baseline.
In conclusion, the researchers said, “We conducted a nationally representative quasi-experimental study of hemodialysis outcomes in a select group of facilities that appear to assign the exact same initial treatment times to all their incident patients irrespective of clinical risk. We find that survival within 2 years is higher among patients who initiate hemodialysis therapy in facilities with ≥4 hours of hemodialysis relative to their counterparts who initiate hemodialysis in facilities that treat all incident patients for 3 hours. This analysis suggests that longer hemodialysis session length may improve outcomes, although further study is needed to establish the degree to which these findings generalize to facilities that prescribe varying hemodialysis session durations.”
- Researchers conducted a retrospective cohort study to compare mortality rates of patients with end-stage renal disease treated in dialysis facilities that used initial session durations of either ≥4 hours or 3 hours for all incident patients.
- At 2 years following initiation of dialysis, there were 8945 deaths in the ≥4-hour group and 15,624 deaths in the 3-hour group. The 2-year mortality rate in the ≥4-hour group was 13.1/100 person-years versus 20.5/100 person-years in the 3-hour group.
- The 2-year adjusted hazard ratio in the ≥4-hour group relative to the 3-hour group was 0.77 (95% confidence interval, 0.70-0.84).