Holliday-Segar maintenance fluids

Daily maintenance volume uses the 100-50-20 rule: 100 mL/kg for the first 10 kg, plus 50 mL/kg for each kg between 10 and 20 kg, plus 20 mL/kg for each kg above 20 kg. The hourly rate uses the derivative 4-2-1 rule: 4 mL/kg/hr for the first 10 kg, 2 mL/kg/hr for each kg between 10 and 20 kg, and 1 mL/kg/hr for each kg above 20 kg. Both rules describe the same water need, but the hourly coefficients are the daily coefficients divided by 24 and rounded (100÷24≈4.17, 50÷24≈2.08, 20÷24≈0.83), so the hourly rate multiplied by 24 does not exactly equal the daily volume. A third figure applies the 100 mL/hour maximum that current guidance states — it takes the smaller of the 4-2-1 rate and 100 mL/hour, and under the 4-2-1 rule it starts to bite at exactly 60 kg. That maximum comes from NICE, Leung 2021, the Royal Children's Hospital Melbourne and the Belgian Be-PIV consensus, NOT from Holliday and Segar, whose function rises without a ceiling. No cap is applied to the daily volume, because guidelines do not agree on one.

Estimate of baseline (basal-state) maintenance water only — not a prescription or an order. It EXCLUDES any pre-existing fluid deficit (dehydration) and ongoing/abnormal losses (fever, vomiting, diarrhoea, drains, third-spacing), which are computed and added separately. There are no interpretation bands: the output is a target volume, not a risk tier. PROVENANCE, STATED FIRST BECAUSE MOST OF THE FOLKLORE AROUND THIS METHOD IS MIS-SOURCED. The 1957 paper was NOT read in full for this implementation. Everything attributed to it here reaches us through the AAP 2018 guideline's direct citation of it, the AAP structured summary, and Chesney's 1998 commentary — three routes that agree. Nothing on this page claims more than that. The 1957 method is denominated in CALORIES, not millilitres; the mL/kg/day form in universal use is a later substitution, and the '4-2-1' hourly rule does not appear in the paper at all — it is a later anaesthetic simplification, and a separate, slightly different one was published by Oh (Anesthesiology 1980;53(4):351). LOWER BOUND, AND WHY IT REJECTS. This score refuses below 4 kg. The familiar 'not intended for neonates younger than about 2 weeks' limitation — which this score itself carried in v1.0.0 — is NOT traceable to 1957: it is traced to a medical-calculator web page carrying a 1998 modification date, from which it propagates into Wikipedia and downstream calculators. The guidelines set the bottom of scope independently and do not agree: ESPNIC 2022 covers term infants from 37 weeks gestation, AAP 2018 starts at 28 days, Leung 2021 at over 28 days, RCH 2026 at one month, and NICE covers term neonates through a separate day-of-life ladder instead. A weight guard cannot implement any of those, because weight does not distinguish the populations: a 3.2 kg term neonate on day 2 needs roughly 70–80 mL/kg/day while a well 3.2 kg two-month-old needs 100 mL/kg/day, a difference of about a quarter at an identical entry. So the behaviour is to require postnatal age where it can be had, reject below 4 kg, and never silently compute; this score does not collect postnatal age, so it takes the strict option and rejects. WHERE THE 4 kg ITSELF COMES FROM, STATED PLAINLY BECAUSE THIS PAGE USED TO OVERSTATE IT. It comes from this project. NO WEIGHT FLOOR FOR THIS METHOD IS CITABLE ANYWHERE — every guideline scope above is written in AGE, and not one of them names a weight below which the method must not be used. The two figures that resemble a floor are not one: ESPNIC's 2024 prescribing figure and RCH's 2026 table both start their bottom band at 3 kg, which is where a table begins, not a rule about what to refuse, and neither states a rationale or says anything about 2.9 kg. Up to v2.0.0 this page presented 4 kg as the source review's recommendation, which gave a number a pedigree it does not have. It is an IMPLEMENTATION CHOICE: deliberate, defensible, and unsourced. The reason behind it is the part that carries weight and it is unchanged — a weight input cannot implement an age rule, so the score refuses across the band where the two populations overlap and where being wrong is worst. A different team could pick 3.5 or 5 and contradict nothing published. What would replace the proxy is collecting postnatal age. 4 kg is a scope floor, not a physiologic threshold and not a published one. The exclusion is well evidenced: in 174 infants of at least 34 weeks gestation on dextrose-containing fluid at a mean 57.2 mL/kg/day, 39% reached sodium at or below 134 and 24% at or below 132 mEq/L, serum sodium fell 0.07 mEq/L per mL/kg of positive fluid balance, and TERM infants fared worse than late preterm ones (31% vs 17% reaching 132 or below) — the opposite of what extrapolating downward from paediatric bands would assume (Chang 2025). Isotonic fluid also raises hypernatraemia risk specifically in neonates (RR 3.74; Amer 2024). Separately, infants aged 1 to 3 months warrant a fluid-TYPE flag rather than a volume change: most tonicity trials recruited from 3 months, and immature kidneys make hypernatraemia and hyperchloraemic acidosis more likely on 0.9% saline, so Leung 2021 statement 5.4 suggests a dextrose-containing balanced solution with lower sodium chloride content and electrolyte monitoring for that band. UPPER BOUND AND THE RATE CAP. 60 kg is the defensible paediatric anchor: it is where RCH's band structure stops and its cap begins, and no source read supports a higher one. The 70 kg anchor that circulates for this method rests on a secondary description of a figure in the 1957 paper that nobody inspected, so it is UNVERIFIED and is not shipped here in any form. The 100 mL/hour rate maximum is the one figure every source states identically — NICE, Leung 2021, RCH 2026 and the Belgian Be-PIV consensus — and under the 4-2-1 rule it binds at exactly 60 kg, so the two anchors coincide arithmetically rather than one being derived from the other. It is a guideline overlay and is never attributable to Holliday and Segar, whose function is monotonic with no ceiling. The 150 kg entry limit is an input-sanity ceiling that asserts nothing clinical. Above the paediatric range the method is replaced rather than extended: body-surface-area forms (300–400 mL/m²/day of insensible loss plus urine output, or 1500 mL/m²/day in total), ideal body weight for the maintenance rate in obesity, or adult guidance at 25–30 mL/kg/day. NICE switches to the BSA form above the 91st weight centile, or in acute kidney injury, known chronic kidney disease, or cancer. Leung 6.1 also states plainly that maintenance fluid should not be prescribed at a rate ABOVE the calculated maintenance rate. THE DAILY CAP, AND WHY THERE ISN'T ONE HERE. There is no ceiling in the original method, and current guidance disagrees with itself across a 500 mL range: 2000 mL/day (ESPNIC group's own 2024 prescribing figure), 2400 mL/day (RCH's January 2026 guideline, stated without citation, and the Belgian Be-PIV consensus), 2500 mL/day (RCH's own fluids calculator page, contradicting its guideline), and 2500 mL for males with 2000 mL for females (NICE, restated by Leung) — while ESPNIC 2022 and AAP 2018 state no cap at all, the AAP explicitly declining to address rate or volume. The widely repeated 2400 mL/day figure is a CITATION ERROR at its point of entry: Be-PIV attributes it to NICE and Leung, both of which were read in full and neither of which states it. It is arithmetically 100 mL/h × 24 h and nothing more. Applied to 70 kg the 1957 arithmetic yields 2500, not 2400. So no daily cap is applied to the volume shown, no single figure is presented as authoritative, and any cap a unit chooses to apply is a guideline convention rather than a derived value. AND THAT ABSENCE IS SETTLED, NOT OUTSTANDING. NO EVIDENCE-BASED DAILY CEILING EXISTS — it is not that a derived maximum sits somewhere unfound, it is that none has ever been derived, which is exactly why four sources give four figures for one question and none of them cites a study that tested a ceiling against anything. Confirmed 2026-08-04 and closed: the disagreement set out above IS the state of the field, and searching it again will not resolve it. WHY THE FORMULA OVERESTIMATES, QUANTIFIED. The 1957 derivation estimated a hospitalised child's caloric expenditure as roughly midway between basal requirement and the requirement of a normally active child. Measured by calorimetry, energy expenditure in acutely ill or postoperative children sits close to basal metabolic rate and averages 50–60 kcal/kg/day (AAP 2018) — against the formula's assumed 100 kcal/kg for the first 10 kg, roughly double. The structural objection is simpler still: the approach was developed from studies of healthy children, and hospitalised children are by definition not healthy (Brossier 2024). This is the quantitative reason current guidance restricts. Accuracy is further reduced in fever, burns, tachypnoea, hypothermia, hyperthyroidism, status epilepticus, and any state with altered ADH physiology. ONE CALCULATOR HERE DELIBERATELY DISAGREES WITH THIS ONE. The pediatric burn resuscitation score reimplements this method and applies it from 0.5 kg, below the 4 kg floor this page enforces. That is intentional, not an oversight: refusing to compute for a burned 3 kg infant would withhold the RESUSCITATION volume too, which is the worse harm. It discloses the same scope limit on its own maintenance output instead of refusing, and a test reads this page's floor directly so the two cannot drift apart silently. RESTRICTION IS NOT HYPONATRAEMIA PROPHYLAXIS. Restriction is defensible for avoiding fluid overload and ESPNIC recommends it on that basis, grading its own volume recommendations C, D and GCP — the weakest in that document — and stating that the amount and duration are uncertain. It is NOT established as a substitute for correct tonicity, and that is no longer an inference drawn from subgroups — one trial randomised the two variables against each other. NEVILLE 2010 (J Pediatr 156(2):313–319, PMID 19818450; abstract read, full text not accessed) allocated 124 postoperative children in a 2 × 2 factorial design: 0.9% against 0.45% saline, CROSSED WITH 100% against 50% of the maintenance rate. The crossing is the whole value of it, because it is what lets an effect be attributed to tonicity or to rate rather than to both at once, which a single restricted arm cannot do. Its conclusion, in its own words: the risk of hyponatremia was decreased by isotonic saline solution 'but not fluid restriction'. NO TRIAL BETWEEN 2016 AND 2026 REPEATED THAT DESIGN — nothing in the last decade re-randomised RATE independently of tonicity, so a 2010 study remains the design of record, and that absence is settled rather than outstanding. Two further sources point the same way: the AAP found the increased hyponatraemia risk from hypotonic fluid PERSISTED in the subgroup given fluid at a restricted rate, and Leung 2021 concludes from the Cochrane review's restricted-rate arms that 0.45% saline under 70% maintenance did not protect against hyponatraemia and that fluid type matters more than fluid rate. None of this contradicts ESPNIC's recommendation to restrict, which rests on fluid overload rather than on sodium; what it forbids is presenting a reduced volume as protection against hyponatraemia. There is currently no reliable way to predict the daily maintenance requirement of a child in acute or critical care, and a causal link between restriction strategies and reduced fluid overload remains to be shown (Brossier 2024). COMPOSITION. The 1957 electrolyte figures are 3 mEq sodium, 2 mEq potassium and 2 mEq chloride PER 100 kcal METABOLISED — not per kilogram. Scaling them per kilogram is wrong, and wrong in a direction that matters at the extremes of weight; this score computes no electrolytes at all, so it cannot make that error, and the figures are recorded only so that any future composition feature starts from the right denominator. Those concentrations were estimated to reflect the composition of human and cow's milk, which is why the original prescription was hypotonic and why no current guideline treats that basis as defensible. The VOLUME rule stands; the 1957 COMPOSITION is superseded. Patients aged 28 days to 18 years receiving maintenance fluid should receive isotonic solutions with appropriate potassium chloride and dextrose (AAP 2018 key action statement 1A, evidence quality A, strength strong; number needed to treat 7.5 to prevent one sodium below 135, 27.8 for one below 130), ideally balanced rather than saline (ESPNIC 2022), though lactate-buffered solutions should be avoided in severe liver dysfunction. Electrolytes and glucose should be checked before starting and at least every 24 hours thereafter. DAILY VERSUS HOURLY. The two outputs disagree slightly by design: the 4-2-1 rule uses rounded coefficients, so hourly × 24 ≠ daily volume — it slightly underestimates in the first two brackets and slightly overestimates above 20 kg.