# Child Dose Check source notes

Reference data updated on 13 October 2026. Licensed source checks were supplied by the Tongy-authorised local chat using signed-in AMH and eMIMS tabs. AMH Children's Dosing was the July 2026 edition; individual PI revision dates remain distinct from the check date. This app compares the entered amount with a displayed reference regimen. The usual reference appears automatically. Schedule, indication and course remain the pharmacist’s checks.

## AMH provenance

All encoded AMH rows now use the directly read AMH Children's Dosing findings, July 2026. The local browser chat checked each represented medicine on 13 October 2026. Source URLs and check dates accompany the rows in `src/data.mjs`. Earlier Memorise Medicine transcriptions were the starting snapshot; they are not the provenance of the current encoded AMH rows. Neonatal, specialist, renal-adjusted and other unrepresented regimens remain outside the quick check.

The site's active JavaScript was inspected, including the functions that actually render product-information panels. The initial data object's MIMS fields are not the values displayed for many drugs. Copying those fields would produce incorrect daily-versus-per-dose comparisons.

## Brand product information checked in eMIMS

| Medicine | Checked brand PI | PI revision | MIMS revision |
| --- | --- | --- | --- |
| Cefalexin | [Otitis media, Cefalexin Sandoz PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=514a5f36-eaad-4338-9539-aae20087190a&type=fullpi) | 2024-07-15 | 2024-09-01 |
| Amoxicillin | [Adult schedule, severe, Amoxil PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=0f52f888-7ced-430e-bc04-a53300fd62c5&type=fullpi) | 2024-05-20 | 2024-07-01 |
| Amoxicillin + clavulanate | [Severe / otitis / lower respiratory, Curam 4:1 PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=670eb06a-da16-42d2-b67c-a53300fddc2d&type=fullpi) | 2026-08-20 | 2026-10-01 |
| Amoxicillin + clavulanate | [Moderate to severe, Augmentin Duo 400 PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=f73dd724-2e79-4690-b114-a53300fd9221&type=fullpi) | 2026-05-15 | 2026-10-01 |
| Flucloxacillin | [2 to 10 years, Flucil oral PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=03618986-7dbf-4225-bb69-a53300fde82f&type=fullpi) | 2025-10-20 | 2025-12-01 |
| Phenoxymethylpenicillin | [Weight-based daily dose, AFT PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=c991f63c-c105-4f99-87e2-a53300fde946&type=fullpi) | 2026-06-03 | 2026-08-01 |
| Cefaclor | [Serious infection / otitis, Sun PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=31019331-9878-4760-963f-b06d00970413&type=fullpi) | 2024-09-20 | 2024-11-01 |
| Erythromycin | [Minor susceptible infection, BD, E-Mycin PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=85ad7009-537a-430c-99ee-a53300fd7bdf&type=fullpi) | 2026-06-03 | 2026-07-01 |
| Azithromycin | [Trachoma / chlamydial conjunctivitis](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=ce64f52a-a756-44fd-9252-a53300fd7d23&type=fullpi) | 2024-05-31 | 2024-07-01 |
| Clarithromycin | [Age 12+, severe dose, Klacid PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=fb929f06-350b-47a3-9e64-a53300fd799d&type=fullpi) | 2026-05-19 | 2026-07-01 |
| Prednisolone | [Croup, PredMix PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=c8a62979-6374-4d3d-b7db-a53300fd815f&type=fullpi) | 2026-08-26 | 2026-10-01 |
| Prednisolone | [Initial child dose, Redipred PI](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=b3737576-6143-48b3-9852-a53300fd8fa7&type=fullpi) | 2021-04-20 | 2021-06-01 |
| Metronidazole | [Flagyl S age/indication table](https://app-emimselite-com.libraryproxy.griffith.edu.au/medicineview?id=978d4d0a-5e35-4197-b72b-a53300fd7c70&type=fullpi) | 2026-03-04 | 2026-04-01 |

PI values are brand-specific. PredMix asthma/croup and Redipred initial-child-dose rows stay separate. Links open the inspected licensed reference through the user's library access. Calculations use saved data, not a live query to either subscription. The footer tracks when reference data changes. Details keeps the check date and individual source revision dates; a CSS deployment does not change the data date.

## Corrections from licensed sources

- Cefalexin empirical UTI is split by age: 1-3 months uses 25 mg/kg/dose QID for 10 days; 3 months to 16 years uses 20 mg/kg/dose TDS, max 750 mg, for 3 days; above 16 years uses the fixed adult UTI reference. The former 12.5 mg/kg QID UTI row was incorrect for these indications.
- Cefalexin Sandoz Table 1 explicitly includes QID and BD child dosing. QID is no longer an assumed comparison. The adult 1 g/day BD restriction conflicts with higher daily totals in the child table and is not encoded as a universal child maximum. The adult 4 g/day reference is not inherited by child otitis dosing. The over-one-year restriction is kept specifically with streptococcal pharyngitis BD.
- AMH prednisolone asthma is 1 mg/kg daily, max 50 mg, for 3-5 days, with specialist advice below 1 year. AMH croup is 1 mg/kg once, repeated after 12-24 hours only if necessary; no numeric maximum is stated in that section. The 1-2 mg/kg daily with usual max 60 mg is initial autoimmune/inflammatory treatment, not a generic asthma/croup range.
- PredMix PI croup remains 1 mg/kg per dose every 8-12 hours for 48 hours. This differs from AMH and is attributed specifically to PredMix. Redipred instead gives an initial child dose of 0.5 mg/kg/day in 3 or 4 divided doses. Escalation and maintenance instructions remain in its note instead of a pooled range.
- AMH flucloxacillin ordinary oral dosing is 12.5-25 mg/kg/dose QID. Usual adult doses of 250-500 mg are not encoded as hard child maxima. The unsupported separate 25 mg/kg severe row was removed. Flucil's exact 10-year boundary remains unresolved.
- Amoxicillin includes separate AOM TDS, BD and high-resistance upper-limit choices. Amoxil oral PI replaces the earlier capsule-brand source. Exactly 20 kg remains unresolved. AMH general dosing is unchanged at 15-25 mg/kg/dose TDS, max 500 mg.
- Amoxicillin/clavulanate suspension ratios remain separate. The 7:1 serious AMH row changes interval to TDS; it does not widen a BD range. The Duo 400 PI cap uses the independently inspected adult Duo Forte 875/125 mg BD recommendation. That cap is not transferred to a 4:1 or TDS row. Curam 4:1's TDS ceiling remains unresolved despite its referenced adult BD tablet schedules.
- Cefaclor's increase-if-necessary AMH wording is an upper-limit-only option. Erythromycin rheumatic recurrence prevention requires documented penicillin allergy. Its conditional AMH BD option below 1 g/day is not encoded as a universal BD row.
- Azithromycin course days, lower ages and infant pertussis treatment remain separate. The old two-day/traveller-diarrhoea grouping was removed. Zithromax throat courses are discrete 10 or 20 mg/kg/day choices, each capped at 500 mg/day. Its child trachoma paragraph has no numeric cap; the adult 1 g reference is not inherited.
- Klacid's weight-based child formula remains separate from its suggested rounded band table. Its age-12 adult transition is explicitly stated. Metronidazole AMH BD/TDS rows are named anaerobic bacterial infection; Flagyl S's fixed age/indication table remains linked rather than inferred from weight.

## Unresolved boundaries and source limits

Broad child bands do not gate routine amount comparisons. There is no age-entry step. Dose-changing age bands and specific formulation/indication eligibility retain their original amber/red assessments and return manual-check guidance immediately. Their known weight formula may be plotted conditionally; fixed-age tables cannot supply an invented weight range. Valid legacy saved ages are checked against the original source bounds and disclosed in Details; no age is inferred from weight or hidden in examples. A routine result without age does not verify patient age applicability. Legacy age context accepts ages below 18 years and excludes neonatal dosing. Written upper-age bands do not always specify endpoint inclusion. The checker conservatively requests review at exactly 12 years for combination-liquid rows, exactly 16 years between cefalexin UTI child/adult rows, and exactly 10 years for Flucil. Those are interface interpretations, not claims that the references explicitly exclude those ages. Infant cefalexin UTI rows are split at 3 months; borderline-age patients require direct source review. Touching or fractional age bands in the Flagyl S and Penicillin V fixed tables remain outside automatic selection.

A maximum absent from a selected section does not mean unlimited dosing. Other indications, adult paragraphs and different formulations do not automatically supply a child maximum. Curam 4:1 stays yellow on a target match pending confirmation of its referenced numerical ceiling. Renal adjustments, including Duo 400's renal/immature-function restrictions, remain a separate clinical check. E-Mycin's BD 1.6 g/day prohibition is treated as schedule applicability when the raw minimum daily dose already exceeds it, rather than clipping that minimum into an invented green target. Its severe doubled 60-100 mg/kg/day range is a separate TDS/QID choice.

## Status rules

Green means the amount fits the displayed row under its reference schedule and supplied applicability details. For routine child bands with no age supplied, age applicability remains a pharmacist check and is identified as unverified in Details. It does not confirm the prescription frequency, indication or course. There is no frequency confirmation control. Ordinary known inputs immediately use the usual row; explicit formulation mismatches and unclear medicine identity still need correction or a picker.

Yellow covers missing details, a dose within 10% outside a stated boundary, an amount matching a different reference schedule or regimen, an upper-limit-only row, and adult/neonatal review. The 10% band is a display rule; it is not clinical permission to round or change a prescription.

This alternate-reference amber fallback applies to the main source comparison. In the optional Indications view, each miniature reports only its own stated indication and schedule. A red miniature does not become amber because another row fits. Each row retains its own maximum and result, and selecting it changes only that source's main reference. Exact targets and upper limits keep their distinct graph semantics. Known wrong-age and wrong-weight rows are excluded. A row with an unknown age can show its known weight formula conditionally in amber; a fixed-age row remains a manual lookup without a graph. All visible graphs share a zero-based scale, but their ranges are never pooled.

Age-sensitive source rows retain unresolved applicability in amber rather than requesting another input. Routine child-scope bounds remain recorded and are checked whenever a legacy saved age is supplied. Below 5 kg, the existing general scope rule returns an infant/neonatal manual check; a displayed child weight formula is conditional. Invalid saved ages suppress graphs and require manual review. The checker accepts measured weights from 0.5 to 150 kg. These are interface scope rules, not age estimation. The Flucil PI's imprecise “2 to 10 years” band is conservatively encoded below 120 months; age 10 needs direct reference review.

Red means the prescription is outside the selected row, or outside all applicable listed rows. It does not establish that every possible indication or specialist regimen is excluded. A stated hard maximum breach is red for an applicable row, even if only slightly over it. A stated usual maximum instead remains an amber above-usual review. Known wrong age/weight scope cannot supply that patient's maximum, and invalid age clears a definitive cap verdict. An unknown age retains an explicit reference-only cap warning with an applicability caveat.

A known per-dose hard maximum breach is red when that row is applicable or its age scope is still unconfirmed. Daily limits are compared under the displayed reference schedule; the actual prescription daily total is not verified. A missing actual age cannot exclude an unresolved age-dependent row. Adult/neonatal scope review does not certify those patient groups' limits. Known incompatible scopes cannot assert a maximum from that row. Adult weight transitions are kept with their individual source rows.

Arithmetic uses unrounded numbers. Display values have up to three decimal places; a positive value below 0.001 displays as `<0.001` rather than zero. Unit-choice amounts and rewritten queries retain a tiny input's original decimal spelling. Comparisons use a tiny floating-point epsilon, not the displayed rounding.

## Entered-unit ranking

The third query term is the amount per administration. A bare value initially uses the unit ranked by measured weight, confirmed medicine/concentration and dose magnitudes from the cached references. This is an unvalidated input heuristic. It supplies neither a calibrated probability nor clinical validation, and cannot establish the writer's intended unit. It does not select an indication, use assessment colors or read the currently selected source row. Actual age and reference choices do not flip the inferred unit.

The ranking uses weight/formulation-eligible cached rows and each encoded schedule, excluding upper-limit-only rows. Each range contributes the geometric mean of its bounds; the combined magnitude is their median on a logarithmic scale. It compares logarithmic distance for the mL and mg interpretations, with a volume preference of 1.5, or 2.25 for prednisolone, reflecting the user's usual volume entry. Rows awaiting actual age remain in this ranking. These choices are product input rules, not clinically validated evidence about intended units.

Both supported interpretations appear as conversion buttons, with the first-ranked unit first and active. `15 pred 10` shows 10 mL = 50 mg and 10 mg = 2 mL. `15 amox250 125` ranks mg and retains both choices; `15 amox250 250` shows only 250 mg = 5 mL because 250 mL exceeds the input scope. The input limits are 200 mL and 10,000 mg. An explicit out-of-scope choice, such as 250 mL, remains visible with a supported alternative for correction. Ranking a unit does not make its amount acceptable, and neither the dose nor reference changes automatically to obtain green.

Recognised PredMix or Redipred brand tokens choose only their own PI rows and remain intact through unit/strength rewrites. Explicit `mg` or `mL` suffixes always win. Tapping a conversion writes an explicit suffix only into the third query term and preserves actual age, source choices, Indications view and keyboard focus. The choices keep stable positions during switching. Bare values stay bare through medicine/strength confirmation; unsupported typed concentrations remain unresolved. Review the displayed unit before interpreting the reference comparison. Indications remains optional and the compact comparison is the default.

## Scope and verification limits

The app covers the ten oral antibiotics in the comparison calculator. Prednisolone oral liquid is included using the separately verified rows above. It is not a complete antimicrobial formulary or indication checker. Renal/hepatic adjustment, allergies, interactions, treatment duration, weekly courses, neonatal dosing and specialist regimens need their own clinical assessment.

Calculations run in the browser. No patient history is stored by the app. URLs can enter browser history and hosting logs; patient identifiers must not be entered. Theme is encoded in the URL.

Earlier independent reviews describe earlier snapshots. The 13 October data uses direct licensed findings supplied by the authorised local browser chat, followed by implementation checks. These checks do not amount to clinical certification.

## Formulation verification

PredMix and Redipred both contain 6.72 mg/mL prednisolone sodium phosphate equivalent to 5 mg/mL prednisolone. Conversion uses the active equivalent; their dosing sections differ and stay separate. Amoxicillin/clavulanate uses the amoxicillin component and retains each suspension ratio. Flagyl S contains 320 mg metronidazole benzoate per 5 mL equivalent to 200 mg metronidazole per 5 mL; conversion uses the 40 mg/mL active equivalent. Injectable, tablet, neonatal and specialist doses are not inferred from the oral-liquid query.