English | 繁體中文
Sodium bicarbonate (NaHCO₃) dosing calculator for metabolic acidosis A clinical decision-support tool integrating 2018–2026 critical-care RCT evidence
Online: https://liangrxdev.github.io/bicarb-dosing-calc/ (interface in Traditional Chinese)
Helps clinical pharmacists and critical-care teams, in ICU metabolic acidosis, to:
- Calculate the NaHCO₃ replacement amount with the deficit formula
- Judge timing of use against current RCT evidence (scenario stratification)
- Convert to the number of 7% ampoules commonly used in Taiwan
- Provide Na⁺ load warnings and a monitoring checklist
⚠️ Disclaimer: This tool is for reference by healthcare professionals only and does not replace clinical judgment. All dose suggestions must be adjusted to actual ABG and electrolyte results. It contains no patient data and makes no diagnostic recommendations.
| Feature | Description |
|---|---|
| Deficit formula | Vd × weight × (target HCO₃⁻ − measured HCO₃⁻); Vd switchable between 0.5 (standard) / 0.7 (severe); the result only indicates the first-dose order of magnitude |
| Clinical scenario stratification | 4 scenarios, each labelled with its evidence level (strongest evidence / not recommended routinely [vasopressor and IHCA scenarios] / calculation only) |
| pH severity stratification | Automatically grades the entered pH (< 7.10 / 7.10–7.20 / 7.20–7.30 / ≥ 7.30), maps it to RCT inclusion ranges and cross-checks the selected scenario |
| Taiwan ampoule conversion | 7% NaHCO₃ (70 mg/mL) 20 mL/ampoule = 16.67 mEq/ampoule; automatically computes the number of ampoules and the exact volume to draw |
| Na⁺ load warning | Flags sodium load when a single dose > 100 mEq (with NaCl equivalent) |
| Monitoring checklist | Before / during / after; AKI stage and vasopressor items appear dynamically by scenario |
| Evidence summary table | Results of 5 trials at a glance, color-tiered |
Required NaHCO₃ (mEq) = Vd × weight (kg) × (target HCO₃⁻ − measured HCO₃⁻)
- Vd coefficient (HCO₃⁻ volume of distribution): standard 0.5; in severe acidosis (pH < 7.10 or HCO₃⁻ < 10) the bicarbonate space grows to 0.6–0.8 (Garella 1973), conservatively taken as 0.7
- Note: this formula's input is (target − measured HCO₃⁻), so the coefficient is the HCO₃⁻ volume of distribution. Do not mix with the base-deficit method: the 0.3 in
NaHCO₃ = 0.3 × kg × base deficitpairs with BE and is not interchangeable with this formula's 0.5 - The coefficient itself is uncertain: the base-deficit method also uses 0.4 (Jung 2026, Curr Opin Crit Care; per Fujii 2019, apparent Vd ≈ 0.4 L/kg 1 h after infusion, varying with condition and infusion rate). The formula only sets the first-dose magnitude; titrate by ABG thereafter
- Note: this formula's input is (target − measured HCO₃⁻), so the coefficient is the HCO₃⁻ volume of distribution. Do not mix with the base-deficit method: the 0.3 in
- Target HCO₃⁻: ≤ 18 mEq/L recommended; do not normalize to 24 (trial target was pH ≥ 7.30)
- Divided dosing: give ½ first, recheck ABG (30–60 min if unstable or on concentrated infusion; otherwise 1–4 h) before deciding the rest
- Daily maximum: ≤ 500 mEq/24h (BICAR-ICU / BICARICU-2 protocol: 4.2% NaHCO₃ 1000 mL/24h)
| Item | Value |
|---|---|
| Concentration | 7% NaHCO₃ = 70 mg/mL = 0.833 mEq/mL |
| Package | 20 mL / ampoule |
| Content per ampoule | 16.67 mEq |
| Dilution to 4.2% (protocol concentration) | 3 parts 7% + 2 parts water for injection |
| Trial | Design | Population | Main results | KRT/RRT |
|---|---|---|---|---|
| BICAR-ICU (Lancet 2018) | Open-label RCT, N=389 | pH≤7.20, general ICU | Composite endpoint NS; AKI stratum Day-28 mortality ↓ (HR 0.59, P=.013) | 52%→35% ↓ |
| BICARICU-2 (JAMA 2025) | Open-label RCT, N=627 | pH≤7.20 + KDIGO AKI 2–3 | 90-day mortality NS (P=.91) | 50%→35% ↓, NNT≈6.5 |
| SODa-BIC (NEJM 2026) | Double-blind RCT, N=498 (mITT) | pH<7.30 + vasopressors | MAKE30 NS (+1.2 pp, −7.1 to 9.4, P=.78); hypokalemia 1.6% vs 0 → not recommended routinely | 20.9%→16.8% (−3.9 pp, CI crosses 0) |
| BIHCA (JAMA 2026) | Double-blind RCT, N=779 | In-hospital cardiac arrest | ROSC NS (RR 1.05, P=.62) | — |
| Chen et al (CCM 2026) | Meta-analysis, N=1111 | 4 RCTs (through 2025/10) | Mortality RR 0.84 (NS); Bayesian 90.4% | RR 0.69, TSA firm |
Clinical takeaways:
- pH ≤ 7.20 + moderate-to-severe AKI (KDIGO 2–3) → strongest evidence for reducing kidney replacement therapy (KRT/RRT)
- In-hospital cardiac arrest (IHCA) → a double-blind RCT showed no difference in ROSC; not recommended routinely
- Mortality: trends favorable across trials but none reached significance
- Pure frontend: HTML / CSS / JavaScript (no framework, no dependencies, no build step)
- Fonts: Noto Sans TC + DM Mono
- CSS: BEM naming
- Deployment: GitHub Pages (root of the main branch)
Just open index.html in a browser; no server needed.
- No user input is collected or transmitted; all calculation happens locally in the browser
- Clinical data are cited from published RCTs and meta-analyses; source years and institutions are shown in the tool
Last updated: 2026-06-15