Dose Normalization

Resolves disparate representations of drug potency into a common base unit for comparison.

• Converts mass units: mcg → mg → g using power-of-10 scaling
• Handles biological potency units: International Units (IU) and USP units
• Normalizes concentration expressions: mg/mL, %, and ratio strengths (1:1000)
• Maps milliequivalents (mEq) to mass for electrolyte drugs

Example: A patient list showing 'Levothyroxine 0.075 mg' and a discharge order for 'Levothyroxine 75 mcg' are normalized to 75 mcg, confirming equivalence.

Translates varied dosing interval expressions into a standardized daily frequency for cumulative dose calculation.

• Maps Latin abbreviations: BID (twice daily), TID (three times daily), QID (four times daily)
• Resolves clock-time schedules: 'Q8H' vs 'three times a day' vs 'every 8 hours'
• Handles weekly/monthly regimens: 'once weekly' normalized to a daily equivalent
• Flags PRN (as needed) medications for exclusion from scheduled cumulative calculations

Example: 'Metformin 500 mg BID' and 'Metformin 500 mg every 12 hours' both resolve to 1000 mg/day.

Disaggregates composite drug products into their individual active ingredients for accurate ingredient-level comparison.

• Splits combination tablets: 'Lisinopril/HCTZ 20/12.5 mg' into two separate active ingredient entries
• Identifies multi-ingredient formulations: resolves brand names like 'Augmentin' to amoxicillin + clavulanate
• Accounts for prodrug conversion: normalizes to the active metabolite mass when clinically relevant
• Links each component to its RxNorm Ingredient (IN) concept for semantic interoperability

Example: A reconciliation engine comparing 'Exforge 10/160 mg' to separate orders for amlodipine 10 mg and valsartan 160 mg identifies them as equivalent.

Computes the cumulative dose over a defined time window to detect unintentional overdosing or sub-therapeutic gaps.

• Calculates total daily dose (TDD): sum of all normalized doses within a 24-hour period
• Detects overlapping orders: identifies when two active orders for the same ingredient cover the same time period
• Computes area under the curve (AUC) approximations for drugs with cumulative toxicity profiles
• Applies duration normalization: converts 'take for 7 days' into a total exposure metric

Example: An active order for 'Warfarin 5 mg daily' and a new order for 'Warfarin 2.5 mg daily' without a discontinuation yields a flagged TDD of 7.5 mg.

Links textually different but clinically identical medication descriptions to a single normalized concept.

• Resolves brand-generic pairs: 'Lasix' and 'furosemide' map to the same RxNorm clinical drug component
• Handles salt vs. base forms: 'metoprolol tartrate' vs 'metoprolol succinate' are distinguished as non-equivalent
• Normalizes abbreviated forms: 'ASA 81 mg' resolves to 'aspirin 81 mg'
• Applies ontology alignment across RxNorm, SNOMED CT, and ATC classification systems

Example: The engine recognizes that 'APAP 500 mg' and 'paracetamol 500 mg' and 'acetaminophen 500 mg' all refer to the same active ingredient.

Applies deterministic rules and probabilistic thresholds to the normalized data to surface clinically significant mismatches.

• Flags absolute differences: normalized dose differs by >10% between pre-admission and admission lists
• Detects therapeutic duplication: two normalized entries map to the same active ingredient with overlapping schedules
• Identifies omission errors: a pre-admission active ingredient has no corresponding normalized entry on the new orders
• Suppresses clinically insignificant variances: ignores rounding differences like 0.125 mg vs 0.12 mg for digoxin

Example: A BPMH entry for 'atorvastatin 40 mg daily' and an admission order for 'atorvastatin 20 mg daily' triggers a dose-reduction discrepancy requiring pharmacist review.

The algorithmic technique of linking brand-name and generic drug products by resolving their chemical constituents to a common base compound. This process prevents duplicate therapy errors by recognizing that Tylenol 500mg and Acetaminophen 500mg represent the same therapeutic entity. In dose normalization, active ingredient matching is the critical first step that aggregates all formulations of a single chemical before calculating total daily exposure.

The capability of an AI system to chronologically sequence clinical events to validate the logical consistency of a medication timeline. For dose normalization, temporal reasoning resolves overlapping prescriptions and calculates cumulative exposure windows:

• Identifies when a dose increase supersedes a prior order
• Detects gaps between discontinuation and new starts
• Computes total active drug load at any point in time
• Prevents double-counting of medications across care transitions

The clinical logic engine that evaluates a patient's estimated glomerular filtration rate (eGFR) against drug monographs to flag medications requiring reduced dosage or discontinuation. Dose normalization must account for renal function because the standardized dose for a patient with impaired clearance differs from the population-normal dose. The system cross-references normalized doses against renal dosing guidelines to verify appropriateness.

A quantitative metric calculated from the total number of concurrent medications, often weighted by anticholinergic burden or sedative load. Dose normalization feeds directly into polypharmacy scoring by providing the accurate, deduplicated count of active ingredients and their cumulative daily doses. This enables stratification of patients at risk for adverse geriatric outcomes such as falls, cognitive impairment, and drug-drug interactions.