Why Dose Units Need to Be Consistent
A longitudinal dose record is only interpretable if its amount and unit are read together. The numerical value 250 could mean 250 micrograms, 250 milligrams, or 250 grams; these are not interchangeable entries. If values in different units are combined as though they shared one scale, dose comparisons and later model inputs can be wrong by factors of 1,000 or more.
The goal is not to erase the source record or to decide that an unusual dose must be an error. Instead, retain the recorded amount and unit, then create a separate normalized amount expressed in a documented canonical unit. A conversion is justified only when the original unit is known and the quantities have compatible dimensions. If those conditions are uncertain, flag the record for investigation rather than silently changing it.
Dimensional Conversion
For mass units, the conversion factor must express the same mass in two units. For example, \(1\ \mathrm{g}=1000\ \mathrm{mg}\), and \(1\ \mathrm{mg}=1000\ \mathrm{\mu g}\). Multiplying by a ratio of equivalent quantities changes the numerical value while preserving the physical amount. Units should be written into the calculation so that unwanted units cancel.
This expression is valid only when the ratio is a correct equivalence for the substance and the units cancel. For a canonical unit of milligrams, a recorded mass in grams can be converted using \(1000\ \mathrm{mg}/1\ \mathrm{g}\). A value already expressed in milligrams uses a factor of \(1\ \mathrm{mg}/1\ \mathrm{mg}\). Keeping the units visible helps reveal an invalid operation, such as multiplying a mass by a volume conversion factor.
A project should select its canonical unit based on the data and intended analysis, document it, and apply it consistently. Milligrams are a practical choice for many mass-based records, but they are not universally appropriate. The unit choice itself does not establish that a recorded value is correct.
Worked Example: Converting Mixed Mass Units to Milligrams
Suppose a hypothetical dataset contains three recorded dose amounts: 0.25 g, 400 mg, and 75,000 micrograms. Convert each to milligrams. Assume the source unit for each record has been verified.
For the first record, use \(1\ \mathrm{g}=1000\ \mathrm{mg}\):
The gram unit cancels, leaving milligrams. The second record is already in the canonical unit: \(400\ \mathrm{mg}\times(1\ \mathrm{mg}/1\ \mathrm{mg})=400\ \mathrm{mg}\). For the third, use \(1\ \mathrm{mg}=1000\ \mathrm{\mu g}\):
The normalized values are therefore 250 mg, 400 mg, and 75 mg. The original values and units should still be retained as recorded; the normalized values are additional derived fields, not replacements.
When Product Strength Is Needed
A record may give a count of dosage units or a volume rather than a mass amount. To derive the mass amount, the product strength must be known and must correspond to the product actually recorded. Dimensional analysis makes the required information explicit: a strength in milligrams per tablet can convert a number of tablets to milligrams, while a strength in milligrams per millilitre can convert a measured volume to milligrams.
The units tablet or millilitre cancel, leaving milligrams. This calculation does not establish that the recorded count or volume was administered correctly. Nor can a product strength be borrowed from a different formulation merely because the drug name is the same.
Worked Example: Converting Tablets to a Mass Amount
In a hypothetical record, the administered count is 2 tablets. The verified product strength is 25 mg per tablet. Derive the amount in milligrams.
Multiply the count by the strength, showing the units:
The derived amount is 50 mg. The calculation depends on both inputs: if the tablet strength were unknown, or if the product identity were ambiguous, the amount could not be obtained from the tablet count alone. Retain the recorded count and the strength information used in the derivation so that the result is auditable.
Worked Example: Checking Equivalent Liquid Records
Two hypothetical records refer to the same formulation. Record A gives 5 mL and a verified strength of 10 mg/mL. Record B gives 0.05 L with that same strength. Convert both amounts to milligrams and assess whether they agree.
For Record A:
First convert Record B's volume to millilitres, using \(1\ \mathrm{L}=1000\ \mathrm{mL}\): \(0.05\ \mathrm{L}\times(1000\ \mathrm{mL}/1\ \mathrm{L})=50\ \mathrm{mL}\). Then apply the product strength:
The derived amounts, 50 mg and 500 mg, differ by a factor of 10. The dimensional calculations are valid, but the records do not agree. This is a discrepancy to investigate: for example, one volume, unit, or product detail may have been entered incorrectly. The calculation alone cannot determine which record is right.
Preserve the Evidence and Check the Scale
A reliable data workflow keeps the raw record intact and records the transformation separately. At minimum, retain the source amount and unit, the derived normalized amount and unit, and the conversion information used. If the value was derived from a product count or volume, retain the corresponding product strength and formulation identifier as well. This allows an auditor to reproduce the arithmetic and helps prevent a later change to the strength table from silently obscuring the original calculation.
Dimensional checks are necessary but not sufficient. A calculation can have units that cancel correctly and still use a wrong source value, an incorrect strength, or a decimal entered in the wrong place. Compare normalized values with the expected range for the specific drug, formulation, and context, and compare records that should represent equivalent quantities. Such checks identify candidates for review; they do not prove which entry is correct.
A useful quality-control sequence is:
Confirm that the amount has an explicit, interpretable unit and that the unit belongs to the recorded value rather than an adjacent field.
For a mass-to-mass conversion, confirm the source unit. For a count or volume, confirm the correct product strength and formulation.
Verify that the source units cancel and that the result has the intended canonical unit.
Check for implausible magnitudes, inconsistent equivalent records, and unexpected product-strength differences. Preserve the evidence and seek source clarification when needed.
What Unit Conversion Cannot Resolve
Some recorded dose units are not mass units. International units (IU), for example, express biological activity, and conversion to mass requires a substance- and product-specific relationship. There is no universal conversion from IU to milligrams. Similarly, a dose recorded in mg/kg is an amount per body mass, not an absolute mass amount; converting it to milligrams requires the relevant body weight and confirmation that this is the intended derived quantity. Do not force either value into a mass-only canonical field without the necessary information and a documented rule.
Product strengths also require care. Two formulations of the same drug can have different strengths, and even an apparently familiar strength should be checked against the product recorded for that event. If a strength is missing or ambiguous, flag the derived amount as unavailable or unresolved under the dataset’s conventions; do not fill it using an unverified assumption.
The main alternatives are to store only source values and convert repeatedly during analysis, or to store a normalized value alongside the source. Retaining only source values avoids premature transformation but makes it easier for different analyses to apply inconsistent conversions. Repeated conversion also makes audits harder. Storing only normalized values is compact, but discards evidence needed to verify the conversion or correct an error. A source-plus-derived design is usually more transparent, provided the transformation and its inputs are documented.
Worked Example: Flagging a Possible Decimal or Unit Error
In a hypothetical check, a verified source record states 0.5 g and the dataset contains a normalized value of 0.5 mg. Determine whether the conversion is consistent.
Using \(1\ \mathrm{g}=1000\ \mathrm{mg}\), the normalized value implied by the source is:
The stored normalized value, 0.5 mg, is 1,000 times smaller than the calculated 500 mg. It does not follow from the stated source amount and unit, so the record fails the conversion check. The appropriate action is to flag it and inspect the raw entry and transformation. If the source actually meant 0.5 mg, the source unit may be wrong; if it meant 0.5 g, the normalized field is wrong. Arithmetic identifies the conflict but cannot decide which source interpretation is true.
Common Mistakes and Practical Limits
- Replacing the source amount with the converted value. Keep the original amount and unit; store the normalized amount as a separate derived value.
- Converting the number but not the unit. A value of 0.25 g is not 0.25 mg. Apply an explicit equivalence factor and verify that the units cancel.
- Using a strength from a different product. A count or volume does not determine a mass amount without the correct formulation-specific strength.
- Treating a flagged discrepancy as proof of an error’s direction. A mismatch can identify a problem but cannot show whether the amount, unit, strength, or derived value is wrong.
- Converting non-mass units as if they were mass. IU and mg/kg require additional substance-specific or participant-specific information before an absolute mass amount can be calculated.
The practical limit is important: normalization improves comparability only when its inputs are reliable. It cannot repair an ambiguous source record, establish what was actually administered, or replace verification of product identity. A defensible dataset makes both the normalized value and the uncertainty in its derivation visible.
Check Your Understanding
Answer each question, then compare your reasoning with the solutions below.
- A verified dose is recorded as 0.08 g. What is its mass in milligrams?
- A hypothetical liquid record gives 3 mL of a formulation with verified strength 20 mg/mL. What amount in milligrams does this imply?
- A record says 400 micrograms, but its normalized value is 400 mg. Does the conversion agree? State the correct normalized amount.
- Why is a tablet count insufficient to calculate an amount in milligrams when product strength is unknown?
- Can 100 IU be converted to milligrams using the general mass-unit conversion factors? Explain.
Solutions. (1) \(0.08\ \mathrm{g}\times(1000\ \mathrm{mg}/1\ \mathrm{g})=80\ \mathrm{mg}\). (2) \(3\ \mathrm{mL}\times(20\ \mathrm{mg}/1\ \mathrm{mL})=60\ \mathrm{mg}\). (3) No. Since \(1000\ \mathrm{\mu g}=1\ \mathrm{mg}\), \(400\ \mathrm{\mu g}=0.4\ \mathrm{mg}\); the proposed value is 1,000 times larger. (4) The amount per tablet is needed so that tablets cancel in the calculation and milligrams remain. (5) No. IU measures activity, and conversion to mass requires a product- or substance-specific relationship; general mass-unit factors do not supply it.