This tutorial introduces the population pharmacokinetic modeling novice to the data format appropriate for most PK model fitting software systems with a focus on intravenous dosing data.
We briefly describe a data format typically used in pharmacokinetics (PK) and pharmacodynamics (PD) modeling, especially relevant to our example intravenous (IV) dosing data. We describe the data form using data items required in NONMEM, which is acceptable in most PK software. More extensive description of data format appropriate for PK software can be found in the manual of each software such as the MONMEM manual\(^{1}\).
Here is an example of data for intravenously administered drugs in a suitable format for PK analysis using most PK software including NONMEM.
id time amt dv rate mdv evid gender
1 1.2 0.00 50.000 NA 0.0 1 1 1
2 1.2 2.05 250.000 NA 0.0 1 1 1
3 1.2 6.25 NA 2.83 NA 0 0 1
4 1.2 20.68 NA 0.41 NA 0 0 1
5 1.2 70.90 NA 0.04 NA 0 0 1
6 1.2 95.25 NA 0.01 NA 0 0 1
7 2.1 0.00 100.000 NA 0.0 1 1 1
8 3.1 1.37 75.000 NA 0.0 1 1 0
9 3.1 20.50 NA 0.70 NA 0 0 0
10 3.1 46.80 NA 0.14 NA 0 0 0
11 3.1 68.83 NA 0.06 NA 0 0 0
12 4.1 0.00 50.000 NA 0.0 1 1 1
13 4.1 0.48 100.000 NA 0.0 1 1 1
14 4.1 1.75 100.000 NA 0.0 1 1 1
15 6.1 29.87 12.100 NA 12.1 1 1 1
16 6.1 37.98 24.000 NA 0.0 1 1 1
17 6.1 38.53 1581.067 NA 24.2 1 1 1
18 6.1 42.27 12.000 NA 0.0 1 1 1
19 6.1 64.97 18.000 NA 0.0 1 1 1
20 6.1 192.83 25.000 NA 0.0 1 1 1
21 6.1 257.37 NA 0.71 NA 0 0 1
This simple example has two types of data rows, dose events (information concerning dosing) and concentration events (information concerning blood concentration levels for the drug of interest). The columns indicate the following information:
id
: unique identifier for each individual in the dataset.time
: time of either dosing or concentration measurement.amt
: dose amount; NA
for concentration events.dv
: drug blood concentration value, which is DV (dependent variable) as NONMEM data item; NA
for a dose event.rate
: rate of drug infusion per unit time. rate=0 for bolus doses.mdv
: missing dependent variable; 1 for indicating that there is no dependent variable (in this case, blood concentration), 0 for dependent variable.evid
: event ID; 1 for indicating dose event (amt & rate for this record will be used for the drug dose information if evid = 1), 0 for observation (or dependent variable if mdv = 0).gender
: an example covariate; each row can have covariates - these can be time varying or fixed within individual.For the first subject (id 1.2) a bolus dose of 50 units is given at time 0 followed by 250 units at time 2.05. The remaining four records for this subject are concentration events. The last subject (id 6.1) has both bolus and infusion dosing events. First, a bolus dose of 25 units is given at time 0 followed by a concentration measurement of 2.28 at time 12.87. The next event is an infusion of 116.4625 units at a rate of 18.15 units per hour (assuming the unit for time is hours) beginning at time 13.70. Infusion time can be calculated by dividing the amount by the rate. For id 6.1, the first infusion time is 116.4625 units/18.15 units per hour = \(6.41\bar{6}\) hours (or 6 hours, 25 minutes).
This PK data form will be generated using Build-PK-IV module (see Build-PK-IV : Simple or Build-PK-IV : Comprehensive ).
Note that additional data items need to be specified depending on dosing schedule or PK/PD models. See PK Data for Oral Dosing for usage of ii
and addl
, which is useful when additional doses (addl
) with the same amount are given with regular interval (ii
), and see MONMEM manual\(^{1}\) for details on cmt
specifying compartment number for dosing or an observation and ss
indicating steady-state dosing.
If you see mistakes or want to suggest changes, please create an issue on the source repository.