**3.2 'Early' versus 'Late' dialysis**

There is conflicting evidence concerning the effect of the early initiation of dialysis on survival. Some retrospective and uncontrolled prospective studies have reported no survival benefits with early dialysis while others have found that early start of dialysis may be harmful (8). After a comprehensive review of the published literature by the National

smooth transition from CKD care to ESRD. Poor planning for initiation of dialysis is a major cause of increased morbidity and mortality. The use of temporary or tunneled dialysis catheters contributes to dialysis mortality by increasing the incidence of sepsis, acting as a stimulus for chronic inflammation, and damaging the central veins, thereby preventing or shortening the survival of more permanent vascular access once created. The chapter will

The appropriate time to initiate dialysis for a patient is not clearly defined. The decision to initiate dialysis in a patient with CKD involves the consideration of subjective and objective parameters by the physician and the patient. Over the past decade a trend of increasing estimated glomerular filtration rate (eGFR) at the initiation of dialysis for treatment of ESRD has been noted in the United States. In 1996, only 19% of patients began dialysis therapy with an eGFR of greater than 10 ml/min/1.73 m2 (denoted as 'early start'), but by 2005 the fraction of early start dialysis patients had risen to 45% (5). It is not known whether early start of dialysis is beneficial, harmful or neutral with respect to the outcome of dialysis treatment for ESRD (5). The timing of initiation of dialysis for ESRD is a matter of clinical judgment guided by values of residual renal function and symptoms and signs present in the patients, including those related to comorbidity. By the time the eGFR falls below 10 ml/min/1.73 m2, most patients require dialysis. However, many patients appear to function quite well until the eGFR approaches 5 ml/min/1.73 m2. As a general rule, patients with diabetes require earlier intervention (eGFR<15 ml/min/1.73 m2) than do those with other etiologies for renal failure (6). Clearly, dialysis must be initiated before the uremic symptoms of peripheral neuropathy, encephalopathy, malnutrition, or serositis (including

 Progressive uremic encephalopathy or neuropathy, with signs such as confusion, asterixis, myoclonus, wrist or foot drop, or, in severe cases, seizures (urgent

A clinically significant bleeding diathesis attributable to uremia (urgent indication)

There is conflicting evidence concerning the effect of the early initiation of dialysis on survival. Some retrospective and uncontrolled prospective studies have reported no survival benefits with early dialysis while others have found that early start of dialysis may be harmful (8). After a comprehensive review of the published literature by the National

 Persistent metabolic disturbances that are refractory to medical therapy; these include hyperkalemia, metabolic acidosis, hypercalcemia, hypocalcemia, and

Hypertension poorly responsive to antihypertensive medications

Table 1. Clinical indications to initiate dialysis in patients with CKD. (7)

discuss in detail regarding type of dialysis access.

**3.1 Indications for initiation of dialysis** 

pericarditis) become evident (See Table 1).

Fluid overload refractory to diuretics

indication)

hyperphosphatemia Persistent nausea and vomiting Evidence of malnutrition

**3.2 'Early' versus 'Late' dialysis** 

Pericarditis or pleuritis (urgent indication)

Kidney Foundation (NKF) workgroup in 1997, they recommended that initiation of dialysis be considered when the arithmetic mean of the urea and creatinine clearances fell below approximately 10.5 ml/min/1.73 m2 except in well-nourished, asymptomatic patients (9). In 1999 Obrador *et al.*, observed that 23% of the US ESRD population, between 1995 and 1997, started dialysis at an eGFR less than 5 ml/min/1.73 m2. They opined that this 'late start' of dialysis needed further examination, including studies of the impact on outcomes and cost of ESRD treatment (10).

In 2006, the NKF work group updated the guidelines for initiation of hemodialysis and stated that 'at CKD Stage 5, when the eGFR is < 15 ml/min/1.73 m2, that nephrologists should evaluate the benefits, risks and disadvantages of beginning renal replacement therapy'. They also suggested that initiation of dialysis therapy before CKD Stage 5 (an eGFR of > 15 ml/min/1.73 m2) may be appropriate in patients who have symptoms believed to be related to both their comorbidities and their level of residual kidney function (11). Only one study has reported the outcomes of patients with CKD who initiated dialysis only after the onset of symptoms due to uremia. In this prospective cohort study of 233 consecutive patients with advanced uremia, 151 were elective starters on dialysis, while 82 initially declined dialysis. Among the initial refusers, 55 percent developed a uremic emergency, while 48 percent were eventually established on maintenance dialysis. In this study, one year mortality was significantly higher among the initial refusers than the elective starters (18 versus 7 percent). However, these results are confounded by lack of randomization and by three deaths among the initial refusers resulting from treatment withdrawal (12).

Additional published studies have not been able to demonstrate any clear-cut survival benefits for early start of dialysis. The only randomized controlled trial that examined mortality and time of dialysis initiation, the IDEAL study (13), found no difference in survival between early or late initiation of dialysis. In this study, 828 patients with progressive CKD and an estimated GFR between 10.0 and 15.0 mL/min/1.73 m2 (as determined by the Cockcroft-Gault equation) were randomly assigned to dialysis initiation when the estimated GFR was either 10 to 14 mL/min/1.73 m2 or 5 to 7 mL/min/1.73 m2. The median time to the initiation of dialysis was 1.8 and 7.4 months in the early and late start groups, respectively. At a median follow-up period of 3.6 years, the authors noted no significant difference in survival (38 and 37 percent mortality, hazard ratio of 1.05 with early initiation, 95% CI of 0.83 to 1.30) as well as no difference in cardiovascular events, infections, or dialysis complications between the late start group and early start group.

However, these results do not imply that the initiation of dialysis can be delayed until the GFR is between 5 to 7 mL/min/1.73 m2 in all patients. The design of the IDEAL study permitted clinicians to initiate dialysis based upon the presence of symptoms due to uremia as well as on the estimated GFR. As a result, 76 percent of patients assigned to the late start arm initiated dialysis when the GFR was much greater than 5 to 7 mL/min/1.73 m2. This resulted in a mean GFR of 9.8 mL/min/1.73 m2 at the start of dialysis for the late start group, which was only 2.2 mL/min/1.73 m2 less than the mean start GFR for the early group (12.0 mL/min/1.73 m2). Thus, approximately 88 percent of all enrolled patients had initiated dialysis with an estimated GFR of approximately 10 mL/min/1.73 m2 or more, either because of symptoms or enrollment in the early dialysis arm (13).

A recent study published in *Canadian Medical Association Journal* examined trends in initiation of hemodialysis within Canada and compared the risk of death between patients with early and late initiation of dialysis (14). Using the Canadian Organ Replacement

Hemodialysis Principles and Controversies 231

The European Best Practice Guidelines for Hemodialysis recommends the standard hemodialysis dose should be delivered as three times per week for 4 hours each session (29). In an attempt to improve outcomes, it was postulated that a higher dialysis dose than commonly provided during conventional dialysis may increase survival among patients

However, this hypothesis was refuted in two large well-designed studies in both



In light of these negative studies, significant attention has turned to alternative dialysis schedules, such as, short-daily HD (SHD), nocturnal HD (NHD), and long, intermittent hemodialysis (LHD). It is suggested that more frequent dialysis may be associated an improvement in health-related quality of life (HRQoL) and with improved survival (33-34). The first successful use of short daily, or "quotidian" hemodialysis was first reported by DePalma in 1969 (35). This approach was based upon the premise that improved patient outcomes, compared with conventional three times per week hemodialysis, would occur with a dialysis schedule that consisted of the same number of hours of dialysis per week but delivered over twice as many sessions. More specifically, it involves five to seven treatments per week, each lasting 1.5 to 2.5 hours. The rationale for short daily hemodialysis is based upon a strategy that is proposed to enhance both dialysis efficiency and hemodynamic stability. With short daily dialysis, shortening the dialysis time while increasing the frequency of dialysis allows more time to be spent dialyzing against higher uremic solute concentration gradients. This enhances the efficiency of solute removal (36). More frequent dialysis allows for less interdialytic fluid accumulation. This is likely to improve hemodynamic stability during dialysis with increased potential for normalizing the extracellular fluid volume. This form of therapy has been associated with significant improvement in serum albumin, calcium phosphate, and volume control in small scale

A recent study, Frequent Hemodialysis Network (FHN) Daily Trial, was a multicenter, randomized trial that included 245 patients assigned to either frequent hemodialysis (six times weekly) or conventional hemodialysis. Two primary composite outcomes were determined at one year, including death or one-year change from baseline in left ventricular (LV) mass as assessed by cardiac resonance imaging, and death or one-year change in physical health as assessed by a RAND health survey. Both composite outcomes showed significant benefit to the frequent-dialysis group compared with the conventional-dialysis group (with hazard ratios of 0.61, 95% CI, 0.46-0.82 for death or change in LV mass; and 0.70, 95% CI, 0.53-0.92, for death or change in physical health) (37). This study also demonstrated benefits in pre-determined secondary outcomes to the frequent dialysis group such as a decrease in LV mass, improved blood pressure control and phosphate balance but not on cognitive performance, depression, serum albumin concentration, or use of

Nocturnal hemodialysis (e.g, long nightly home hemodialysis) was introduced as a potentially more desirable alternative to conventional dialysis, since it provides superior dialysis based upon dose, duration, and frequency (38). This can be accomplished because it is performed during nightly sleep, an otherwise unproductive time (39). The late Robert

of a thrice weekly regimen failed to decrease patient mortality (31).

undergoing renal replacement therapies (30).

hemodialysis and peritoneal dialysis patients:

greater than a weekly Kt/V of 1.7 (32)

studies. However, no mortality data is available.

erythropoiesis-stimulating agents.

Registry from 2001 to 2007, the investigators identified a cohort of 25,910 patients 18 years or older who began hemodialysis. Dialysis was defined as beginning early if the eGFR exceeded 10.5 mL/minute/1.73 m2. Mean eGFR at initiation of dialysis increased from 9.3 mL/minute/1.73 m2 in 2001 to 10.2 mL/minute/1.73 m2 in 2007 (*P* < .001). During the same period, the proportion of early dialysis initiations increased from 28% (95% confidence interval [CI], 27% - 30%) to 36% (95% CI, 34% - 37%). Among those starting dialysis early, mean GFR at initiation was 15.5 mL/minute/1.73 m2 vs 7.1 mL/minute/1.73 m2 among those who started dialysis late. For early vs late initiation of dialysis, the unadjusted hazard ratio (HR) for death was 1.48 (95% CI, 1.43 - 1.54). This suggests that early initation is associated with higher mortality. After adjustment for demographic factors, serum albumin, primary cause of endstage renal disease, type of vascular access, comorbid conditions, late referral, and transplant status, the hazard ratio for death decreased to 1.18 (95% CI, 1.13 - 1.23). Difference in mortality per 1000 patient-years between starting dialysis early vs late decreased after 1 year of followup but persisted and began increasing again after 24 months of follow-up, with significant differences at 6, 12, 30, and 36 months. (14)

There may be two additional advantages to early dialysis: control of hypertension and increased dietary intake. Reversal of volume overload with dialysis often leads to a reduction in blood pressure, which is typically volume-dependent in CKD. Perhaps more important, patients on dialysis patients require at least 1 g/kg of protein per day to replace dialysis losses and maintain nitrogen balance. Thus, early institution of dialysis can allow a more liberal diet in terms of both food and fluid.

The overall conclusion of these trials largely supports current practice that dialysis initiation should be based upon clinical factors rather than the estimated GFR alone. Patients with progressive CKD require close follow up, early nephrology referral, and adequate advance dialysis planning (including the presence of a functioning peritoneal or vascular access and referral for transplantation). Among patients with progressive CKD, clinicians must be vigilant for the presence of symptoms and/or signs of uremia and dialysis should be initiated in the patient with these symptoms.
