Represent Bolus and Infusion Doses

Why Dose Representation Matters

A dose record is not just an amount. It must also describe when the drug entered the body and over what period it was administered. A bolus and an infusion can have the same total amount but supply that amount differently over time. If their records are represented as though they were the same event, the analysis receives the wrong dosing history.

This tutorial focuses on encoding bolus administrations and infusions in analysis-ready dose-event records. It assumes dose amounts have already been expressed in a consistent unit, as established in the earlier material on dose-unit normalization. Dose times must use the same time basis as the corresponding sample records; the detailed handling of clock times and time zones is a separate data-preparation task.

Fields That Describe the Administration

An analysis-ready dose record should preserve enough information to distinguish the input type and reconstruct its timing. A general representation includes a subject identifier, an event type or route, a dose amount with units, a start time, and—when the administration is an infusion—a duration or an end time. A rate may also be recorded directly or calculated from amount and duration. The precise column names depend on the analysis system; the meanings and units must remain unambiguous.

For a bolus, record the amount and the time at which it was given, and identify it as a bolus. A zero duration can be used if the data structure requires a duration field, but it must mean “instantaneous bolus” rather than “an infusion whose duration is unknown.” Do not calculate a finite infusion rate by dividing a bolus amount by zero.

For an infusion, record the start time and either the duration or the end time, along with the total amount and/or delivery rate as supported by the source record and analysis format. If both amount and duration are known, the average rate over that interval is

where \(D\) is the total dose amount, \(T\) is the infusion duration, and \(R\) is the average delivery rate. Conversely, if rate and duration are known for the complete administration, the total amount is

The units provide a basic check on the calculation: amount divided by time gives amount per time, such as mg/h. Keep the amount, rate, and time units explicit. A value of 200 without units is not a sufficiently defined rate.

Representing the Input Over Time

The distinction can be stated mathematically without committing to a particular software format. Let \(t_0\) be the administration start time. A bolus supplies amount \(D\) at \(t_0\); it is a discrete input at that time, not a constant rate across a time interval. An infusion with constant rate \(R\), starting at \(t_0\) and lasting \(T\), supplies drug during the interval from \(t_0\) to \(t_0+T\). Its delivered amount is the rate integrated over that interval:

This equality justifies calculating a rate from total amount and duration when both refer to the same complete infusion and the rate is treated as constant. If the source gives a varying rate, the total amount is instead the integral of that rate over time; a single average rate does not preserve the detailed rate pattern.

Translating Source Records Carefully

Source documents may describe an administration using different combinations of fields. Translate what is explicitly recorded, and calculate only quantities that follow from those records. For example, if a record gives total amount and start and end times, calculate duration and, when an average constant rate is appropriate, calculate rate. If it gives rate and the complete start-to-end interval, calculate total amount. Preserve the original recorded values where the data process allows, so a derived quantity can be distinguished from a directly documented one.

A volume-based pump record may require a concentration to convert delivered volume into drug amount. For example, if a hypothetical solution contains 2 mg/mL and the participant receives 250 mL, the amount is \(2\ \text{mg/mL}\times250\ \text{mL}=500\) mg. If that volume is infused over 2.5 h, the average rate in amount units is \(500\ \text{mg}/2.5\ \text{h}=200\) mg/h. A pump rate in mL/h is not itself a drug delivery rate in mg/h unless the solution concentration is also known.

Do not treat a prescribed amount as proof of the amount actually administered. A planned dose and a documented administration are different facts. If the source provides only a planned amount, label or store it accordingly rather than silently representing it as received. Likewise, if the source lacks a duration or an end time for an infusion, do not invent one merely to fill a field. Record what is known and identify what remains unresolved for data review.

Choosing and Checking a Representation

Two common infusion representations are amount plus duration and rate plus duration. When the rate is constant, either pair determines the other quantity through \(D=RT\). A record containing amount, duration, and rate can be useful when it is clear which values came from the source and which were calculated. The benefit is traceability; the risk is that copied or independently edited values may not describe the same event. Do not make a field appear more certain than its source.

After translation, check that the event type is explicit, amount and rate units are present, and start time and duration have clear units and share a time basis with the rest of the analysis data. Confirm that calculated values are marked as derived and that the source supports the assumptions used. A zero duration should not ambiguously stand for both a bolus and missing infusion timing.

Common Mistakes and Limitations

• Encoding an infusion as a bolus. Putting the full amount at the infusion start time erases the time over which the drug was delivered. Retain the infusion duration or equivalent timing information.
• Calling a rate an amount. A value in mg/h is not a dose amount in mg. Keep the units attached to each field and use \(D=RT\) only when the interval and rate support that calculation.
• Using zero duration to mean “unknown.” Zero duration represents an instantaneous event only when the record explicitly identifies a bolus. Unknown timing must remain distinguishable from a bolus.
• Inferring actual administration from a prescription. A prescribed amount does not establish that the participant received it. Represent the source status accurately.
• Converting volume to amount without concentration. A volume such as 250 mL cannot be converted to mg unless the drug concentration is known.

These representations also have limits. A single average rate summarizes a constant-rate infusion well, but it does not encode a changing rate within the interval. If the source documents rate changes or only an incomplete administration, one amount-and-duration pair may not capture the event adequately. Preserve sufficient source detail and flag the need for review rather than forcing a simplified record that implies unsupported information.