Home PK intro R NONMEM Calculators Data checker

NM Datasets

‹ Back

NONMEM datasets

NONMEM requires data in a specific format with some mandatory items. Here is an example of a NONMEM compatible database:




General rules dataset:
Apart from the required items some general rules apply to the dataset.
  • Decimal seperator has to be a '.' and cannot be a ',' (pay attention to language settings)
  • No empty cells are allowed (use e.g. '.' instead)
  • Specific date and time format (like shown in the above dataset)
  • Time has to be in chronological order
  • Censor symbol (to specify which rows should not be read, like the header)
  • Data in CSV format

The mandatory fields are:
  • ID: unique subject identification
  • TIME or DATE TIME: chronological order of dosages and observations
  • DOSE (of AMT): administered dose
  • DV: observed concentration (dependend variable)
  • EVID: event identification
    • 0 = observation
    • 1 = dose
    • 2 = other
    • 3 = reset event (reset compartments)
    • 4 = reset-and-dose (reset system and administer dose; e.g. new patient)
  • MDV: concentration (DV) available (MDV = missing dependend variable)
  • CMT: compartment of dose or observation (e.g. 1 depot, 2 central compartment)
 Semi-required depending on the data:
  • DATE: If calendar dates and clock times (containing ':') are used. When calendar dates are used specify as DATE=DROP. DATE is month day year; DAT1 is day month year; DAT2 is year month day; DAT3 is year day month.
  • RATE: Infusion rate (0 = bolus, >0 = rate, -1 of -2 zero-order bolus)
    RATE for infusions (>0) is amt per hour. E.g. AMT=30 RATE=15 (dose is administered over 2 hours).
  • SS: Steady-state applies (in combination with II)
  • II: Dose interval in case of steady-state (required with SS)
  • ADDL: Additional doses (repeat 'x' times at interval II).

Covariates:
Use columns for the covariates like patients' weight, comedication, lab results, etc. Indicate missing covariates for example using '-99' or '99999'. If covariates are used the value cannot be empty.


Infusions:
There are multiple ways to define infusions in your dataset. Infusion can have a duration (time of infusion), it can be a bolus infusion, continuous infusion, etc.
  • Bolus infusion: The most simple form is an iv administration, where it just has a dose (AMT) specified and injected into the central compartment. If the dose is injected at another place then where the concentrations are measured, specify the correct compartment (CMT), absorption and maybe fraction.
  • Infusion over time: If the infusion is administered over an specific time period, the RATE should be specified in conjunction with the AMT. RATE is the same unit as AMT over time. So if AMT is mg and time in hours, RATE is specified as mg/h. NOTE: RATE is NOT time, it is unit per time.
  • Continuous infusion: Has a TIME (when sample is taken during continuous infusion), RATE (infusion rate e.g. mg/h), SS (SS=1 indicating steady-state). NOTE: the infusion stops at the time the sample is taken. Hence, it needs to be repeated in case of multiple samples during continuous infusion.
  • Multiple infusions at steady-state: If an infusion is repeated at an fixed interval it can be specified as SS=1 (steady-state) with interval II (e.g. II=12 is every twelve hours), AMT and RATE. This creates prior infusions (to -inf) using the specified interval, dose and rate. Multiple infusions can run in parallel, for example to have different dosages at different interval.