RE: how to include inconsistent infusion rate into the model

Hi Tomi, I haven't done infusion rates that vary with individual. However, you can model this with between subject variability described on your infusion rates. If you know what the rate should be, then you can fix that and allow each individual to vary. Matt. >From the NONMEM help | | | RATE DATA ITEM | |_________________________________________________________________| MEANING: Dose Rate (RATE) data item for PREDPP CONTEXT: $INPUT record and NONMEM data set USAGE: $INPUT ... RATE ... DISCUSSION: RATE labels PREDPP's dose rate (RATE) data item. The dose rate data item is optional. In a dose or reset-dose event record it can take four kinds of values. 0 The dose is an instantaneous bolus dose. >0 The dose is an infusion, and the value of RATE gives its rate. If this is not a steady-state infusion, the AMT data item must be positive also. If this is a steady-state dose event record, both AMT and RATE may be 0; (See ss dose). -1 The dose is a zero-order bolus dose, i.e., an infusion, whose rate is modeled in the PK routine. As an example, suppose the compartment being dosed is numbered 1 (e.g., the CMT data item contains 1, or it contains 0 and compartment 1 is the default compartment for doses). The $PK abbreviated code must define a value for R1, the rate parameter for compartment 1, e.g., R1=THETA(4)+ETA(4) -2 The dose is a zero-order bolus dose, i.e., an infusion, whose duration is modeled in the PK routine. As an example, suppose the compartment being dosed is numbered 1 as above. The $PK abbreviated code must define a value for D1, the duration parame- ter for compartment 1, e.g., D1=THETA(4)+ETA(4) RATE must be 0 for observation, other-type, and reset event records.