Home Hemodialysis with NxStage

2.1 What frequency buys — the FHN Daily Trial

The FHN Daily Trial randomized 245 patients to 6×/week in-center HD versus conventional 3×/week for 12 months [1]. The frequent arm averaged 5.2 sessions/week and reached a weekly standardized Kt/V of 3.54 ± 0.56 versus 2.49 ± 0.27 — separation you can only get from frequency, not from intensifying a thrice-weekly schedule.

Both co-primary composite outcomes favored frequent HD:

LV mass fell by 16.4 g with frequent HD versus 2.6 g conventional (P<0.001). Blood-pressure and phosphorus control improved. Intradialytic hypotension was less common (10.9% versus 13.6% of monitored sessions, P=0.04). The cost was access — vascular-access interventions rose (HR 1.71, 1.08–2.73). There was no effect on cognition, self-reported depression, serum albumin, or ESA dose [1].

At a median 3.6 years, the mortality separation reached significance: 20/125 (16%) deaths with frequent HD versus 34/120 (28%) conventional, HR 0.54 (0.31–0.93) [2]. Frequent HD also reversed adverse ventricular remodeling, with the largest effect in patients who had minimal residual urine [5].

2.2 The nocturnal signal — read with caution

The FHN Nocturnal Trial randomized only 87 patients to 6×/week home nocturnal HD versus conventional [3]. Despite delivering 1.82-fold higher weekly stdKt/V and 2.45-fold more treatment time, neither co-primary outcome was significant (death/LV mass HR 0.68; death/physical-health HR 0.91). The trial was underpowered. Hyperphosphatemia and hypertension improved.

Extended follow-up then produced a higher mortality hazard in the nocturnal arm — HR 3.88 (1.27–11.79) [4]. The authors caution against over-reading it: the comparator was largely home-based conventional HD with an unusually low event rate (0.03 deaths/patient-year), the sample was tiny, and post-trial prescriptions changed heavily. It is hypothesis-generating, not a verdict. The honest summary: frequency’s benefit is best-established for short-daily; the nocturnal mortality signal is unexplained and contested.

2.3 Comparative and registry data

• Daily HHD versus peritoneal dialysis (Weinhandl, USRDS, 4,201 matched pairs): 20% lower all-cause mortality (HR 0.80, 0.73–0.87), 8% lower hospitalization, 37% lower technique failure [6]. Caveat — among patients starting home therapy within 6 months of ESRD onset, mortality was similar (HR 0.95).
• ANZDATA (52,097 patients, 1998–2017): home HD mortality ran below facility HD throughout, even as uptake shifted toward older, sicker patients [7].
• Technique-failure penalty (ANZDATA, Semple): after home-HD failure and return in-center, adjusted mortality was markedly higher (HR 3.93 at 0–30 days, still 2.29 beyond 90 days) [8]. That is an argument for supporting patients well on home therapy, not a property of home HD itself.
• Extended-hours meta-analysis (Hull, 476 patients, 6 trials): no proven QoL or mortality benefit, no excess access events, most trials at high risk of bias [9]. The ACTIVE Dialysis RCT (200 patients) found no difference in quality of life or LV mass with extended hours, but lower phosphate and potassium, higher hemoglobin, and fewer BP and binder medications [10].

2.4 Patient-centered outcomes

The FREEDOM cohort reported the outcomes patients actually feel. Post-dialysis recovery time fell from 476 to 63 minutes [11]. Depression scores improved (BDI 11.2→7.8; proportion with depressive symptoms 41%→27%) [11], and restless-legs severity dropped [12]. Antihypertensive burden fell from 1.7 to 1.0 agents, with the proportion on none rising 21%→47% [13]. Frequent HD also shortens recovery by roughly an hour versus conventional [14]. Culleton’s nocturnal RCT discontinued or reduced antihypertensives in 16/26 patients and stopped binders in 19/26 [15].

3.1 Low-volume dialysate and the flow-fraction concept

Conventional in-center HD floods the dialyzer with 500–800 mL/min of dialysate. NxStage does the opposite and exploits saturation. When dialysate flow is low relative to blood flow, spent dialysate approaches equilibrium with blood urea, so nearly every liter is fully used. Clearance is the product of saturation (the dialysate-to-plasma ratio, D/P) and dialysate flow rate [25].

Flow Fraction (FF) is the NxStage-specific term for the ratio:

FF = dialysate (effluent) flow ÷ blood flow

Two numbers matter, and they are not the same:

• Physiologic ideal: dialysate saturation approaches 100% at FF 25–33% (Qb roughly 3× Qd) [25]. Manufacturer guidance says saturation exceeds 90% when blood flow is roughly 3× dialysate flow.
• Operator-practical: UpToDate reports that experienced programs “typically use a flow fraction of 40 to 50 percent to make efficient use of the dialysate and achieve the target Kt/V without substantially increasing the time on dialysis” [25].

The tension is time versus dialysate. A lower FF wrings more clearance out of each liter but takes longer; a higher FF is faster but leaves each liter less used. Above roughly FF 0.50 the trade turns bad — saturation drops below 90% and you spend dialysate (bags, PureFlow time, cost) on clearance you do not bank. Peer-reviewed work confirms the physiology: at low FF, urea clearance approximately equals dialysate flow rate [16]. Practical range: FF 0.40–0.50, with 0.30–0.35 reserved for when time is not the constraint.

Two consequences worth internalizing:

1. A fixed flow fraction holds delivered Kt/V steady even if blood flow wobbles intradialytically — a stability advantage at home [25].
2. Because the ratio is fixed, ultrafiltration adds to the total fluid leaving the dialyzer and therefore adds time — roughly 6–10 minutes per extra liter of UF [25]. In some patients that becomes its own incentive to limit fluid intake.

3.2 PureFlow SAK versus bagged dialysate

Practical rule: anything above approximately 30 L/treatment is a PureFlow job. Forty liters on bags is eight bags to hang daily — cumbersome and costly. A single batch cannot exceed 60 L, so very large patients on less-frequent schedules must combine sources.

4.1 Frequency and time

• 3.5–4×/week mitigates the long interdialytic gap and lowers ultrafiltration rate.
• 5–6×/week adds the survival, BP, and UFR benefit seen in FHN and FREEDOM [1,2,11].
• Minimum 12 hours/week; 15 or more preferred. Phosphate benefit scales with total weekly time — nocturnal (25 or more h/week) is the strongest phosphate lever.
• Short-daily sessions run approximately 2.5–4 hours; nocturnal uses lower blood flow (200–300 mL/min) over longer times.

4.2 Dialysate volume and body size

The clearance floor is fixed by physiology:

Minimum dialysate volume per treatment = Total Body Water (V) × target per-session spKt/V (at 100% saturation)

Estimate V by the Watson equation (Therapy Handbook coefficients: women V = 0.42 × weight, men V = 0.50 × weight; a bedside proxy is approximately 50% of body weight). A field shortcut used by NxStage prescribers and echoed by OpenEvidence: daily dialysate volume ≈ V × 0.4. Worked example — a 99-kg patient with Watson V of 49 L receives 49 × 0.4 ≈ 19.6 L, and since Qd is fixed, time follows (volume ÷ Qd). Because real saturation is approximately 90%, not 100%, run slightly above the floor. Note that delivered spKt/V typically runs approximately 0.05 above prescribed, because the Watson equation modestly overestimates V [25].

NxStage Dosing-Calculator volume ranges (Therapy Handbook; assumes approximately 1.5 L UF, FF 0.30–0.40, sessions <3 h), for a stdKt/V target of 2.1:

*>60 L per treatment requires PureFlow plus bags. A stdKt/V target of 2.3 shifts every cell higher. The signal from this table is simple: at 5–6×/week, most patients are dosed at 20–30 L. Frequency does the work. Body size and lower frequency drive volume up.

4.3 Blood flow, dialysate flow, flow fraction

Set Qb high relative to Qd (typically 400–450 mL/min) to keep FF in the efficient range. Time ≈ (dialysate volume + UF volume) ÷ Qd.

4.4 Ultrafiltration rate — the mortality-linked setting

A central advantage of daily HD: the same weekly fluid, spread over 5–6 sessions, keeps per-session UF rate low. That matters.

• UFR > 13 mL/kg/h → all-cause mortality HR 1.59, CV mortality HR 1.71 (Flythe, HEMO cohort) [28]. Confirmed in 118,394 patients — >13 HR 1.31, >10 HR 1.22 (Assimon) [29]. Risk starts climbing above approximately 10 mL/kg/h.
• Manufacturer guidance is more conservative still: UFR ≤ 8 mL/kg/h suggests lower mortality risk (frequent HHD averages approximately 6.0 versus approximately 9.4 mL/kg/h in-center).

Target: keep UFR < 13 mL/kg/h, ideally < 10. Achieve it by adding frequency or time and holding interdialytic weight gain ≤ 3 kg — not by pulling harder per session [30].

5.1 Potassium — the arrhythmia lever

Low-potassium dialysate is a modifiable sudden-cardiac-death risk.

• Dialysate K < 2 mEq/L was associated with sudden cardiac arrest, independent of pre-dialysis serum potassium (Pun, 43,200 patients; 4.5 SCA events per 100,000 treatments) [31].
• In DOPPS, dialysate K < 3 mEq/L was associated with sudden death; short treatment time, large UF volume, and low Kt/V compounded it [32]. Sudden death is roughly a quarter of all dialysis deaths [33].

Frequent-HD nuance — a defensible extrapolation, not a frequent-HD RCT: daily dialysis produces lower pre-dialysis potassium and a smaller serum-to-dialysate gradient, so an aggressive 1–2 mEq/L bath is usually unnecessary and potentially hazardous. Default most daily-HD patients to K 2–3. Reserve K 1 for documented, persistent hyperkalemia. Never combine the lowest-K bath with rapid UF or a QT-prolonging drug. The Handbook targets a post-dialysis serum K > 3.1 mEq/L. Select by pre- and post-dialysis potassium.

5.2 Calcium

Frequent dialysis removes calcium in smaller, recurrent increments, which is why the NxStage standard bath (3.0–3.5 mEq/L) sits higher than the low-calcium in-center baths. For nocturnal or otherwise intensive schedules, the higher 3.5 mEq/L (1.75 mmol/L) is often indicated to offset greater cumulative calcium removal [35]. Titrate by serum calcium and PTH. Remember that low-calcium dialysate independently predisposes to intradialytic hypotension and arrhythmia [31,36].

5.3 Buffer (lactate)

NxStage dialysate is lactate-buffered — lactate converts roughly 1:1 to bicarbonate. Select by pre-dialysis serum bicarbonate: use 45 mEq/L for HCO₃⁻ ≤ 20, and 35–40 mEq/L for HCO₃⁻ ≥ 24, targeting a mid-week pre-treatment bicarbonate of 22–26 mEq/L. Modeling shows dialysate lactate concentration and weekly dialysate volume are the primary determinants of pre-dialysis serum tCO₂; on nocturnal schedules 45 mEq/L can over-correct, and reducing volume limits the rise [18]. Prefer bicarbonate-based solutions if lactate metabolism is impaired (hepatic failure, lactic acidosis).

6.1 Laboratory panel and cadence

The monthly cadence for adequacy and nutrition is standard operational practice. KDIGO minimums for MBD and anemia are less frequent, and the interval should scale to the size of any abnormality.

6.2 Measuring delivered dose at home

• Draw pre-dialysis BUN and a correctly sampled post-dialysis BUN on one representative monthly session. Compute spKt/V (Daugirdas second-generation equation), then convert to stdKt/V including UF and residual clearance [23].
• Post-BUN sampling technique matters. Use the slow-blood-flow/stop-pump method — reduce Qb to approximately 50–100 mL/min for approximately 15 seconds, or stop the pump 10–20 seconds before drawing — so access recirculation does not falsely lower the post-BUN and overstate the dose.
• Always convert to stdKt/V before judging adequacy. A per-session spKt/V of 0.5 looks inadequate against the thrice-weekly floor of 1.2, but it is on target at 5–6×/week (see §8.2).

6.3 Water and dialysate quality (PureFlow)

Two standards are in play. Cite the one in force locally.

CMS requires bacteriological and endotoxin testing at least quarterly for home systems, per manufacturer labeling. PureFlow is validated to ultrapure grade (bench <0.1 CFU/mL, <0.03 EU/mL for up to 12 weeks); because the system is preconfigured and validated, NxStage labeling does not mandate routine per-batch dialysate culturing. Confirm the exact interval against current FDA labeling and unit policy.

8.1 First principle: dose is a floor, not a goal

There are no RCTs powered on hard outcomes that set a small-solute clearance goal in home HD. And stdKt/V did not predict mortality, hospitalization, or transfer among 2,373 home-HD patients [25,26]. Clear the Kt/V floor, then judge success on clinical endpoints — volume and BP, phosphorus, acid-base, electrolytes, anemia, and nutrition.

8.2 Delivered-dose targets by frequency

KDOQI: for schedules other than thrice-weekly, minimum delivered stdKt/V 2.1, target 2.3 [21]. Because per-session dose is small by design, the spKt/V needed to reach stdKt/V of approximately 2.1 depends on frequency [25]:

For orientation, FHN Daily delivered stdKt/V 3.54 at 6×/week [1]. You cannot meaningfully raise weekly dose inside a 3×/week schedule — frequency and time are the levers, not per-session intensity. Residual kidney function counts toward total stdKt/V and legitimately lowers the required dialytic dose; the variable-target model formalizes this [27]. These targets rest on established kinetics — the standardized Kt/V model [19] and the equivalent renal urea clearance concept [20] — and on the KDOQI supporting systematic review, which found that more-than-thrice-weekly and extended-length HD did not improve hard clinical outcomes versus conventional therapy [22], consistent with the FHN solute-clearance analyses [24].

8.3 The multidimensional check — signs to change

9.1 Mineral-bone — the phosphate signature, and its reversal

Frequent HD is the most effective phosphate control available (FHN) [37]:

• Daily arm: pre-dialysis phosphorus −0.46 mg/dL, binder dose −1.35 g/day; 73% needed no binders at 12 months versus 8% on 3×/week.
• Nocturnal arm: pre-dialysis phosphorus −1.24 mg/dL; 42% required phosphate added to the dialysate to prevent hypophosphatemia.
• Calcium and PTH were essentially unchanged — manage those with dialysate calcium (§5.2), vitamin D analogs, and calcimimetics, not by changing frequency.

Convergent data: nocturnal binder discontinuation near 75% [38], and an every-other-day nocturnal OL-HDF series saw binder use fall 77%→3% with 58% needing dialysate phosphate supplementation [39]. Action: as binders come off on intensive or nocturnal schedules, monitor for hypophosphatemia and add phosphate to the bath per unit protocol.

9.2 Anemia and nutrition — set honest expectations

RCTs show no anemia or ESA benefit from frequency [1,15]. Some intensive or convective cohorts report lower ESA and iron needs, but that is not RCT-proven [39]. Do not expect ESA savings from short-daily HD. Albumin does not improve with dose [40] — nutrition is a separate workstream.

9.3 Troubleshooting and when to return in-center

• Recurrent circuit clotting → check heparin dosing (typical NxStage: 1,000–2,000 U bolus, then 500–1,500 U/h; consider syringe pump or LMWH for long or nocturnal runs), access flow, low-Qb states.
• Pressure/air alarms → cannulation position, line kinks, access recirculation or stenosis.
• Bring a home patient back in-center for: access failure needing intervention, recurrent failed cannulation, catheter-related bacteremia, refractory volume or BP not fixable by prescription, recurrent hemodynamic instability without a partner present, persistently inadequate delivered dose with symptoms, or care-partner burnout and safety concerns — a leading real-world cause of technique failure.