A mechanism-driven selection and prescription framework for hemodynamically vulnerable HD patients.
Iso-UF, sodium ramping, and UF ramping all act on the same four-axis cascade. Selecting the right tool starts with naming which axis is failing in this patient.
Every intradialytic-hypotension (IDH) event is the visible end of an unfavourable arithmetic: instantaneous fluid removal exceeds the patient's instantaneous ability to compensate. The three techniques in this guide are different mechanical answers to the same physiologic question — how do we keep effective circulating volume and vascular tone intact while net body water falls?2
Fluid leaves the intravascular compartment at the ultrafiltration rate (UFR). It is replaced from the interstitium at the plasma-refilling rate, which is finite and falls as the session progresses (interstitial hydrostatic pressure drops, oncotic pressure rises, lymphatic return saturates). When UFR exceeds refilling, relative blood volume falls, cardiac filling falls, and IDH follows.2 This is the dominant mechanism behind weight-based UFR thresholds (≈10–13 mL/kg/hr).14,15
Diffusive removal of urea and other small osmolytes during conventional HD lowers plasma osmolality during the session. Water shifts intracellularly (impairing refilling) and the osmotic drive to vasoconstriction is blunted. This is the specific axis iso-UF and sodium profiling were designed to support: iso-UF avoids diffusive solute loss entirely during the UF block;5 Na ramping replaces lost tonicity early when the UFR is highest.6,9
Adequate refilling and adequate tonicity still require the vasculature to recruit compensatory venoconstriction and shift blood out of the splanchnic bed. Long-standing diabetic autonomic neuropathy, elderly age, polypharmacy with antihypertensives, and severe diastolic dysfunction all blunt this response.3 A patient who cannot vasoconstrict will not be rescued by any single hemodynamic technique — but iso-UF's preserved peripheral vascular resistance (PVR) is one of the few in-session interventions that addresses this axis.4
Diastolic dysfunction, LVH, severe aortic stenosis, and low-output states are all preload-dependent. As preload falls during UF, these hearts cannot augment output. The Pirkle and KDIGO datasets agree: cardiac stress is a function of average UFR, not its distribution. Profiling alone does not protect a stunned myocardium if mean UFR is too high.13,14
Iso-UF protects axes 2 and 3 (preserved osmolality → preserved PVR). Na ramping protects axis 2 (delivered tonicity). UF profiling protects axis 1 (UFR matched to falling refill curve). None protects axis 4 — that requires lowering the mean UFR (more time, more frequency, or lower IDWG).
Repeated IDH is not a per-session inconvenience. Each event delivers a microcirculatory hit to the myocardium (Burton's "myocardial stunning"), cerebral white matter, gut mucosa (with translocation/endotoxemia signals), the vascular access (with thrombosis risk), and any residual nephrons.18 A patient with ten IDH events per month is accumulating end-organ injury at a rate the dialysis prescription itself can modify. This is the population in which these techniques earn their place.
Convective fluid removal without diffusion. Preserves plasma osmolality and peripheral vascular resistance — at the cost of zero clearance during the UF block.
Iso-UF is fluid removal by convection with dialysate flow off or with no diffusive gradient. Solute clearance during the UF phase is essentially nil. It is most commonly delivered as sequential dialysis: a discrete iso-UF block (front-loaded, when refill reserve is highest) is followed by a conventional HD block that delivers the diffusive dose.
Because no urea or small osmolytes are removed during the iso-UF block, plasma osmolality is preserved. Two consequences follow:
It is not "the same UF, with less stress." It is a different hemodynamic event: the tonicity-preservation pathway is closed in conventional HD+UF and open in iso-UF. This is the clinical justification for choosing it over simply slowing UFR.
There is no diffusive clearance during the UF block. Total session Kt/V is the HD-block Kt/V only. If treatment time is not extended, the diffusive dose must be compensated by higher blood-flow, dialysate-flow, or membrane KoA — and the realistic ceiling on these is finite. State the adequacy trade-off explicitly in the order (§7 worksheet).
A tonicity-replacement strategy. Powerful when designed for net-zero diffusive Na — harmful when not.
Sodium ramping (a subset of sodium profiling) means setting the dialysate Na high at session start — typically 150–160 mEq/L — and stepping or linearly decreasing it toward ≈140 mEq/L by session end.12 Profile shapes include:
Early high dialysate Na delivers a positive diffusive Na flux into plasma during the steepest UF interval. The plasma tonicity that would otherwise fall (from urea/osmolyte removal) is supported, the inward water shift is blunted, refilling is preserved, and the osmotic component of vasoconstrictor tone is maintained.9,10 The descending limb aims to return the net diffusive Na balance toward zero by session end.
If the profile is not engineered for net-zero diffusive Na, the patient leaves the session salt-loaded. The downstream cascade is predictable and well-documented:
This is the KDIGO Controversies Conference harm signal: dialysate Na individualization carries a meaningful prescribing-error risk and an indirect signal of harm when imprecisely applied. Avoid sodium loading during dialysis is the conference's explicit caution.1
The Song et al. (JASN 2005) data and the broader Na-balance literature show that profiles engineered for net-zero diffusive Na transfer can preserve the hemodynamic benefit while minimizing the IDWG penalty.8 The operative variable is not "start Na" — it is the time-averaged dialysate Na relative to the patient's pre-dialysis serum Na.
Time-averaged dialysate Na ≈ pre-HD plasma Na. A profile that starts at 155 mEq/L and ends at 137 mEq/L over a 4-hr session has a time-averaged Na of ≈146 mEq/L — and is net-positive Na transfer in a patient with pre-HD Na of 138. Either lower the time-averaged Na (shift the whole profile down), or shorten the high-Na phase, until the gradient closes.
Time-varying UFR within a session. Reshapes the UFR curve to match the falling refill curve — but does not lower the average dose.
UF profiling delivers a time-varying UFR within a single session, in contrast to constant UFR. Common shapes:
The instantaneous plasma-refilling rate is highest at the start of the session and falls progressively as the interstitium empties. A descending UF profile matches the UFR to that curve: remove the most water when refill capacity is greatest, and protect the late-session window where refilling has collapsed and IDH events cluster.2
The Flythe crossover RCT is the most important piece of evidence to represent accurately. UF profiling did reduce intradialytic lightheadedness and reduced post-dialysis hypervolemia — but it did not reduce treatment-induced cardiac stress.13 The greatest documented hemodynamic benefit appears when UF profiling is combined with sodium profiling.9,11
UF profiling reshapes the UFR curve. It does not reduce the average UFR. A patient whose mean UFR exceeds their safe ceiling cannot be rescued by profiling — they need longer time, increased frequency, or a dry-weight reassessment. Profiling that masks an unsafe average is harm, not care.
Profiling is about distribution; the average still matters. Anchor to the literature: the ≤13 mL/kg/hr threshold (and the more conservative 10 mL/kg/hr inflection signal) marks the population-level IDH risk transition, reaffirmed by 2025 reviews placing UF and IDWG as the dominant modifiable drivers of IDH.14,15,18
However — and this is the critical caveat the guide must carry — chasing UFR limits in isolation can cause under-removal and volume expansion. The 2018 review (PMID 29885084) and KDIGO both warn against treating any single UFR cutoff as an absolute rule when the alternative is leaving the patient wet.16 The right framing: the UFR limit is an alarm to reconsider time/frequency, not a license to abandon volume control.
Cool dialysate and dry-weight/time optimization are higher-evidence first moves. The ramping/iso-UF toolkit is layered onto the genuinely refractory or fragile subset.
| Intervention | Mechanism | Primary indication | Evidence | Principal harm / limitation | Adequacy impact | Resource / equipment |
|---|---|---|---|---|---|---|
| Iso-UF Sequential UF→HD | Preserves plasma osmolality & PVR during the UF phase. | Osmolar-collapse–driven IDH, large per-session solute shifts. | Small mechanistic RCTs + KDIGO framing. | Zero clearance in UF phase → adequacy trade-off; longer session. | Negative if HD block not compensated. | Any modern HD machine. |
| Na ramping (balance-neutral) | Replaces lost tonicity early; supports refilling & vascular tone. | IDH-prone with tolerable IDWG, refractory to first-line measures. | Multiple small RCTs / crossover; 2024 combined RCT. | Sodium loading → IDWG, interdialytic HTN, masked volume overload (KDIGO). | Neutral if net-zero Na. | Programmable Na on machine. |
| UF profiling Descending / A/D | Matches UFR to falling refill curve. | Late-session IDH clustering with acceptable mean UFR. | Crossover RCTs (symptom reduction; cardiac stress unchanged). | Masks unsafe mean UFR if used as cover. | Neutral. | Programmable UFR profile. |
| Cool dialysate 35.5°C | Sympathetic activation; thermal vasoconstriction. | IDH-prone; co-first-line non-pharmacologic. | Cochrane (2019) — consistent IDH reduction.19 | Patient thermal discomfort. | Neutral. | Any HD machine. |
| BVM / biofeedback UFC | Real-time refill-guided UFR adjustment. | Frequent, unpredictable IDH where staffing reactive control is limited. | Crossover comparative (2025).17 | Cost; equipment availability; mixed clinical signal. | Neutral. | BVM-capable machine + sensor. |
| Treatment time / frequency increase | Lowers required mean UFR per session. | Anyone with mean UFR > safe ceiling. | Strong observational; some RCT (FHN). | Patient acceptance; chair-time logistics; cost. | Improves. | Operational (chair time). |
| Dry-weight reassessment | Corrects systematic over-prescription of UF. | Suspected wrong-set dry weight (most IDH starts here). | Strong — BIA-supported.20 | Requires repeated bedside assessment; BIA cost. | Neutral. | Clinical / BIA / lung US. |
| Antihypertensive timing review | Reduces pre-HD vasodilatation. | Any IDH-prone patient on multiple antihypertensives. | Pragmatic. | Misses interdialytic BP control if too liberal. | Neutral. | Operational. |
| Midodrine | α-1 vasoconstriction. | Refractory IDH despite the above. | Small studies; KDIGO Conference: insufficient.1 | Supine HTN; cost; QID dosing. | Neutral. | Pharmacy. |
In a resource-aware unit, the rational order is: 1. dry-weight reassessment + treatment-time review · 2. cool dialysate (high-evidence, zero cost) · 3. antihypertensive timing and IDWG counselling · 4. UF profiling for documented late-session IDH · 5. sodium-balance-neutral ramping if IDWG is tolerable · 6. iso-UF / sequential dialysis for the osmolar-collapse phenotype · 7. BVM-UFC if available and the above fails. Pharmacologic measures (midodrine) are layered on the residual refractory tail, not used early as a shortcut.
The clinician must be able to identify the subset of patients these methods will have the greatest impact on — and the subset in whom they will cause harm.
| Clinical driver | Preferred technique | Rationale | Caveat |
|---|---|---|---|
| Osmolar-collapse–driven IDH; large solute shifts (e.g. urea reduction >75%, hyponatremic pre-HD Na). | Iso-UF (sequential) | Preserves tonicity & PVR during the UF block. | Adequacy trade-off; longer total time. |
| Late-session IDH clustering with adequate early tolerance; mean UFR within safe ceiling. | Descending UF profile | Matches UFR to falling refill curve. | Average UFR must still be safe. |
| IDH-prone with tolerable IDWG, refractory to cool dialysate & UF profiling alone. | Na-balance-neutral ramping ± UF profile | Tonicity replacement + refill optimization. | Strict net-zero Na; not for high-IDWG patients. |
| Frequent unpredictable IDH; BVM-capable machine available. | BVM-UFC (biofeedback) | Real-time refill-guided UF rate adjustment. | Cost; evidence mixed; not substitute for selection. |
| Any IDH-prone patient, low resource setting. | Cool dialysate first | High-evidence, near-zero cost. | Thermal discomfort in some. |
Order-sheet templates with embedded safety guardrails. Transcribe into the HD prescription; do not initial off-template.
The order sheet should record the patient's pre-HD plasma Na, the mean UFR (mL/kg/hr) check, and the reassessment date as required fields. These three values prevent the three most common errors: profile-blind Na prescribing, distribution-without-dose UF profiling, and indefinitely continued ramping after the precipitant has resolved.
These are temporizing tools. Every prescription needs a written exit.
| Metric | Why | Cadence | Action threshold |
|---|---|---|---|
| Per-session SBP nadir & IDH events | Primary efficacy endpoint of any prescription in this guide. | Every session. | No reduction over 6 sessions → de-escalate or change tool. |
| Pre-HD BP trend | Detects creeping volume overload from Na loading. | Every session; review trend at 4-session intervals. | Rise >10 mmHg SBP over 4 sessions → stop Na ramping; reassess dry weight. |
| IDWG trend | Earliest signal of sodium loading. | Every session. | Rise >0.5 kg above baseline → stop Na ramping. |
| Post-HD orthostatics | Captures residual hemodynamic vulnerability after the technique is added. | Every 4 sessions. | Persistent orthostatic SBP fall >20 mmHg → mean UFR review; reduce dose. |
| Achieved vs target dry weight | Confirms profiling is enabling — not masking — adequate volume control. | Monthly clinical + 3-monthly BIA if available. | Drift upward → reassess dry weight, not just the prescription. |
| Residual kidney function (urine output) | RKF is a prognostic asset; IDH and excessive UF erode it. | Monthly. | Falling RKF → escalate stability strategy and reduce UF stress. |
| Achieved Kt/V (single-pool) | Iso-UF risk: under-dialysis if HD block does not compensate. | Monthly. | Kt/V < 1.2 → extend HD block or move to longer session. |
Cross-link to the KDIGO 2020 Controversies Conference framework already used in the unit: avoid sodium loading, individualize on the basis of pre-HD plasma Na, and pair every intradialytic stabilization technique with structured volume reassessment.1
The five sentences that, said out loud at handover, prevent the five most common errors.
Sodium ramping is a free win — it makes IDH go away.
Truth: Non-neutral ramping buys intradialytic stability at the cost of chronic volume burden — higher IDWG, interdialytic hypertension, masked dry-weight drift. The KDIGO Conference explicitly cautions against sodium loading during dialysis.1
UF profiling fixes a high average UFR.
Truth: Profiling reshapes the curve; it does not lower the area under it. A patient whose mean UFR exceeds their safe ceiling needs longer time, more frequency, or a corrected dry weight — not a profile.13
Iso-UF improves dialysis clearance.
Truth: Iso-UF delivers zero clearance during the UF phase. Total session Kt/V is the HD-phase Kt/V alone. The adequacy trade-off is real and must be planned for.
These techniques reduce mortality.
Truth: Unproven. The honest claim is symptom and IDH reduction in selected patients. Use them for that, not as life-prolonging interventions.1
The 13 mL/kg/hr UFR limit is an absolute rule.
Truth: Chasing UFR cutoffs in isolation can drive under-removal and volume expansion. The threshold is a signal to reconsider time/frequency, not a license to send the patient home wet.16
Internal Medicine · Nephrology · Nutrition · Philippines · PRC 0105184
Last reviewed:
Educational. Decision support for licensed clinicians. Individualize to patient and local dialysis-unit resources.
Scan and save
