Survival, hospitalization outcomes similar between biweekly, thrice-weekly dialysis schedules

A twice-weekly haemodialysis schedule does not appear to increase death and hospitalization rates at 2 years compared with a thrice-weekly schedule, although some patients on the biweekly schedule may require medication dose adjustments, as reported in the prospective multicentre nonrandomized TATH* trial.

In the intention-to-treat analysis, the primary outcome of survival at 2 years did not significantly differ between the haemodialysis frequency arms. But looking at the causes of death, sudden cardiac deaths occurred more frequently in the twice-weekly arm (53.3 percent vs 38.7 percent), while infection was more common in the thrice-weekly arm (38.7 percent vs 26.7 percent), according to first study author Dr Mabel Aoun from the Université Saint-Joseph in Beirut, Lebanon.

Likewise, no significant differences in the secondary outcomes were observed, Aoun added. At 2 years, the rate of uncontrolled hypertension (defined as >160/90 mm Hg in the last three sessions) was 20.7 percent in the thrice-weekly arm and 18.4 percent in the twice weekly arm (p-value range, 0.442–0.852), while the cumulative number of hospital admissions was 8 and 5 (p-value range, 0.515–0.947), respectively. However, the cumulative erythropoietin (EPO) dose tended to be higher in the twice-weekly arm (720,000 vs 500,000 units; p-value range, 0.515–0.947). [ERA 2024, abstract 1809]

The p-values, according to Aoun, were derived from multivariable logistic, linear, and Poisson regression models, wherein baseline imbalances in patient characteristics were addressed by including propensity scores and baseline residual diuresis as covariates.

Guidelines recommend three 4-hour sessions of haemodialysis per week. But Aoun pointed out that many patients worldwide receive treatment twice weekly, not always as part of an incremental strategy, but due to various reasons including patient preference, travel limitations, or transportation costs.

The findings provide evidence that survival and hospitalization outcomes do not differ significantly by haemodialysis frequency, she said.

TATH trial

TATH involved 204 consecutive patients initiating chronic haemodialysis, of which 132 (mean age 68 years, 53.8 percent male) were in the thrice-weekly arm and 72 (mean age 64 years, 54.9 percent male) were in the twice-weekly arm. None were terminally ill. Baseline characteristics were balanced between the haemodialysis frequency arms, including comorbidities such as diabetes (58.1 percent vs 50.7 percent) and coronary artery disease (48.5 percent vs 34.8 percent). However, significantly more patients in the twice-weekly arm were employed or had jobs at baseline (41 percent vs 25.2 percent; p=0.041).

“Unfortunately, we had a low recruitment rate over 4 years (2018–2021) due to the pandemic and did not reach the target of 806 patients,” Aoun said, adding that TATH had to be terminated prematurely.

All patients started haemodialysis at an eGFR of 6 mL/min/1.73m². There were no significant differences in the baseline levels of serum potassium (4.5 vs 4.6 mmoL/L), haemoglobin (9.6 vs 9.4 g/dL), and serum albumin (33 vs 34 g/L), but there was a trend toward higher residual diuresis in the twice-weekly arm (500 [475-1,000] vs 500 [500-1,000] mL; p=0.075).

Regarding medications, patients on twice-weekly haemodialysis used more antihypertensive drugs at baseline (2 vs 2.6 p=0.005). No other differences were observed.

Total prescribed dialysis hours per week was 12 in the thrice-weekly arm and 8 in the twice-weekly arm (p<0.001). Baseline dialysate flow was significantly lower in the twice-weekly arm (500 mL/min: 54.5 percent vs 87.3 percent; p<0.001), as was blood flow (329 vs 292 mL/min; p=0.001). Dialysate calcium and potassium concentrations were also lower in the twice-weekly arm (p<0.001 for both).

During follow-up, the twice-weekly arm tended to have higher serum potassium levels (1 month: 4.6 vs 4.5 mmoL/L; 3 months: 5.1 vs 4.8 mmoL/L; 1 year: 5.2 vs 4.8 mmoL/L) but lower ferritin concentrations (3 weeks: 217 vs 357 ng/mL; 1 year: 348 vs 495 ng/mL) and transferrin saturation (3 weeks: 19 percent vs percent; 1 year: 24 percent vs 29 percent). No difference was observed in haemoglobin levels.

Eight percent of patients were lost to follow-up, and cross-over occurred in the two arms, with four patients switching to a twice-weekly schedule at 2 years due to economic crisis and nine switching to a thrice-weekly schedule at 3 months probably because of loss of diuresis.

*Twice against thrice-weekly haemodialysis