Why Infusion Fields Must Agree
An infusion is not specified completely by a dose amount alone. Its timing and delivery rate affect when drug enters the system, so a usable event record needs a clear start time and a duration or rate consistent with the amount. A record that says “600 mg” might represent a rapid infusion or a slow one; those are different exposure histories even though the total amount is the same.
The practical task is to prepare the available fields without assuming that every field is correct. If amount, duration, and rate are all recorded, check their agreement. If one is missing, derive it only when the other two have compatible meanings and units. Keep the source values distinguishable from calculated values, and investigate material disagreements instead of silently overwriting one field.
Meaning, Units, and Assumptions
For example, if \(A\) is measured in milligrams and \(D\) in hours, then \(R\) must be in milligrams per hour. The product \((\mathrm{mg/h})(\mathrm{h})\) has units of milligrams, as required. Before calculating, convert quantities to a common amount unit and a common time unit. A numerical calculation with mismatched units can look plausible while being wrong by a factor of 60, 1,000, or more.
The equation \(A=RD\) describes a constant-rate infusion: the rate is the same throughout the stated duration, and the amount refers to what is delivered over that interval. If the rate changes over time, the total amount is instead the integral of the changing rate. In that setting, \(A/D\) is the average rate over the interval, not necessarily the rate at any particular time. Do not treat a recorded rate as constant unless the field definition or source supports that interpretation.
The event should ordinarily have a valid start time and a positive duration. A zero duration cannot be used to calculate a rate by division. A negative duration is invalid for an infusion and should be investigated; although division by a negative number can produce a finite mathematical result, that result does not make the event valid. A negative rate or amount is also a reason to check coding, units, and source data rather than to accept the record at face value.
A start time is not interchangeable with an end time or a sample time. When both start time and duration are present, their implied end can be calculated as a useful check. If the recorded time is missing, a duration and rate may still determine the amount, but the infusion cannot be located adequately in time. The details of time origins and conventions should be consistent with the dataset’s established time basis.
Deriving and Checking Infusion Quantities
Under the constant-rate assumption, any two of \(A\), \(D\), and \(R\) determine the third when the units are compatible and the required divisor is nonzero. In particular, a duration must be positive to represent an infusion. A derived field should be labelled as derived, with its inputs and units retained, rather than presented as if it had been measured independently.
The justification follows directly from the definition of constant rate. Over a small interval of length \(dt\), a constant rate \(R\) delivers amount \(R\,dt\). Summing those amounts across the infusion gives the integral below. Because \(R\) is constant, it can be taken outside the integral, leaving the rate multiplied by the elapsed duration.
The algebra also explains when derivation is not permitted: calculating \(R=A/D\) requires \(D\ne 0\), and calculating \(D=A/R\) requires \(R\ne 0\). For a valid positive-duration infusion, a positive amount and positive constant rate are the usual case. If a field violates those expected conditions, first determine whether it is a genuine special case, a unit or sign error, or an event that should not have been encoded as an infusion.
Worked Example: Deriving Rate From Amount and Duration
A hypothetical infusion record gives an amount of 900 mg over 1.5 h, beginning at 08:15. Assume the amount is the amount delivered during that interval and the rate is constant. Find the rate and the implied end time.
The duration is positive, and the amount and duration units are explicit. Apply \(R=A/D\):
The implied end time is 1.5 h after 08:15. Since 1.5 h is 1 h 30 min, the end time is 09:45. Thus the internally consistent fields are a 900 mg amount, a 1.5 h duration, and a 600 mg/h constant rate, starting at 08:15. This calculation does not establish that the source record is clinically correct; it shows what rate follows from its stated amount and duration under the stated assumption.
Worked Example: Deriving Duration After Unit Conversion
A hypothetical source lists an amount of 1.2 g and a constant rate of 300 mg/h. Find the duration. The amount and rate use different amount units, so convert before dividing.
Since \(1\ \mathrm{g}=1{,}000\ \mathrm{mg}\), the amount is \(1.2\times1{,}000=1{,}200\) mg. The rate is positive, so \(D=A/R\) is defined:
The units reduce as \(\mathrm{mg}/(\mathrm{mg/h})=\mathrm{h}\). The derived duration is 4 h. Dividing the unconverted numerical values, \(1.2/300\), would not produce a meaningful duration because the amount units differ.
Checking Agreement Within a Tolerance
When all three fields are present, calculate the amount implied by rate and duration, \(RD\), and compare it with the recorded amount \(A\). A discrepancy can arise from rounding, unit conversion, transcription, or genuinely incompatible fields. Exact equality is often too strict for rounded source values, so define a tolerance before reviewing records. There is no universal tolerance suitable for every dataset; it should reflect the precision and purpose of the source fields.
One practical rule combines an absolute tolerance \(\tau_{\mathrm{abs}}\), in amount units, with a relative tolerance \(\tau_{\mathrm{rel}}\), which is dimensionless. For a positive recorded amount, the fields pass this consistency check when:
The left side is the absolute amount discrepancy. The right side allows a fixed small discrepancy plus one proportional to the recorded amount. State both tolerance values and their units in the data-preparation specification. A passing check means only that the fields agree within the chosen rule; it does not prove that the source amount, duration, or rate is true.
Worked Example: Applying an Explicit Tolerance
A hypothetical record gives \(A=750\) mg, \(D=2.0\) h, and \(R=370\) mg/h. For this example, suppose the project has specified \(\tau_{\mathrm{abs}}=5\) mg and \(\tau_{\mathrm{rel}}=0.02\). Determine whether the record passes.
First calculate the amount implied by the rate and duration:
The discrepancy is \(\lvert750-740\rvert=10\) mg. The allowed discrepancy is:
Because \(10\ \mathrm{mg}\leq20\ \mathrm{mg}\), the fields pass this specified check. The rate-duration product is 740 mg, not exactly 750 mg; the conclusion is only that the 10 mg difference is within this project’s 20 mg tolerance. A different documented tolerance could lead to a different pass/fail decision.
Choosing Which Fields to Store
The best pair of directly sourced quantities depends on what the source actually reports and on the requirements of the downstream model or data format. Preserve the original values and identify any derived quantity and its calculation. If all three are supplied and disagree beyond tolerance, do not choose one to overwrite the others without a documented resolution rule.
| Available fields | Possible derived field | Key condition or caution |
|---|---|---|
| Amount and duration | Rate \(R=A/D\) | Duration must be positive; units must align. |
| Amount and positive rate | Duration \(D=A/R\) | Rate must be nonzero; the result must be a valid positive duration. |
| Duration and rate | Amount \(A=RD\) | Assume a constant rate over the duration and compatible units. |
| Amount, duration, and rate | Consistency check | Compare \(A\) with \(RD\) using a stated tolerance. |
A common mistake is to treat amount, duration, and rate as interchangeable descriptions and to calculate a missing value without checking what the source fields mean. For example, a rate may be a nominal setting rather than an observed delivery rate, or an amount may refer to a different interval. The arithmetic can be correct while the interpretation is not. Confirm that all fields describe the same infusion event and interval before applying the relationship.
A second mistake is to regard any finite quotient as a valid infusion quantity. With \(A=4\) mg and \(D=-2\) h, \(A/D=-2\) mg/h is mathematically finite, but a negative duration is not a valid infusion duration. By contrast, \(D=0\) makes \(A/D\) undefined. The correction is to check validity conditions as well as algebra: investigate a negative duration, and do not derive a rate by division when duration is zero.
Other useful diagnostics include checking for missing or implausibly large durations, unexpected signs, unit changes within a data source, and an implied end time inconsistent with the event’s recorded timing. These checks flag records for review; they do not identify the correct field automatically. A duration-rate pair can be used to calculate an amount, while an amount-duration pair can be used to calculate a rate, but neither choice resolves which input is wrong when the source fields conflict.
The relationship also has limits. If the delivery rate varies, the product \(RD\) is valid only when \(R\) denotes the average rate over the full interval, or when the rate is constant. If the source amount does not refer to the amount delivered during the stated duration, the fields cannot be reconciled by this formula. Record the assumption and the provenance of derived values so that reviewers can distinguish a documented calculation from a source measurement.
Check that amount, duration, rate, and start time refer to the same infusion, and convert amount and time units to compatible forms.
Confirm that duration is positive for a valid infusion; if deriving duration from amount and rate, confirm that the rate is nonzero and the resulting duration is valid.
Use \(A=RD\) under the constant-rate assumption. Label calculated values, and apply a documented tolerance if all three fields are present.
Retain the source values and send failed checks or invalid fields for review rather than silently replacing a value.
Check Your Understanding
Work through each question, then compare your reasoning with the solutions below.
- A hypothetical infusion delivers 480 mg over 0.8 h at a constant rate. What rate does this imply?
- A source lists 600 mg and 150 mg/h. What duration follows, and what validity condition should be checked?
- A record gives 500 mg, 2 h, and 240 mg/h. Using \(\tau_{\mathrm{abs}}=2\) mg and \(\tau_{\mathrm{rel}}=0.01\), does it pass the stated tolerance rule?
- Why should a negative recorded duration be investigated even though dividing an amount by it produces a finite number?
Solutions. (1) \(R=A/D=480\ \mathrm{mg}/0.8\ \mathrm{h}=600\ \mathrm{mg/h}\). The duration is positive and the units are compatible. (2) \(D=A/R=600\ \mathrm{mg}/(150\ \mathrm{mg/h})=4\ \mathrm{h}\). Check that the rate is nonzero and that the resulting duration is positive and consistent with the event. (3) The implied amount is \(RD=(240\ \mathrm{mg/h})(2\ \mathrm{h})=480\) mg, so the discrepancy is \(\lvert500-480\rvert=20\) mg. The allowed difference is \(2\ \mathrm{mg}+0.01(500\ \mathrm{mg})=7\) mg. Since \(20\ \mathrm{mg}>7\ \mathrm{mg}\), it fails. (4) A negative duration is not valid for an infusion. A finite quotient is only an arithmetic result; it does not validate the inputs or make the event meaningful.