In most instances, the short answer is NO….unless you like protocol deviations that have no use and no bearing on the analysis of pharmacokinetic data from the trial.
Clinical studies involving PK/PD sample collection (e.g., blood) for pharmacokinetic analysis often have sampling windows around the pre-defined collections times specified in the clinical protocol (called nominal collection timepoints). Clients often ask our Pharmacokineticists “what window should we use for collection of PK samples” around nominal timepoints. For example, if the clinical protocol calls for collection of blood at 1 hour post-dose (the nominal timepoint), should samples be collected within a +/- 5 minute time window around that 1 hour timepoint?
From a pharmacokinetic analysis perspective, sampling windows around nominal timepoints are useless. It is much more important that sites record the actual time of pharmacokinetic sample collection than determining if they were in an arbitrary window for collection. When conducting pharmacokinetic analysis (such as noncompartmental analysis in WinNonlin) Pharmacokineticists use the actual time of each sample collection compared to the actual time of dosing to determine elapsed time from administration of study drug. This is the crucial information that is needed to conduct a proper pharmacokinetic analysis.
Sampling windows are more of a clinical conduct issue and have no bearing on the pharmacokinetic analysis that our Pharmacokineticists conduct at Nuventra. We recommend not placing a window around collection of samples for pharmacokinetic analysis and instead focus on ensuring that the actual time of sample collection and the actual dosing time are recorded in the CRF. We also recommend the inclusion of language in the protocol stating that collecting PK samples at times other than nominal timepoints will not qualify as a protocol violation or deviation.
Phase 1 clinical research units (CRUs) are very good at collecting pharmacokinetic samples from healthy volunteers and most often hit every collection timepoint at the exact time specified by the clinical protocol. Phase 2 and 3 investigational sites, and sites working with patients that are critically ill, may need a little more assistance at collecting PK samples at the right time. Nevertheless, by not including a sampling window in the protocol the Sponsor is dictating to the Investigator, CRU, and/or investigational site that pharmacokinetic samples need to be collected at the exact nominal time noted in the protocol. If they are unable to collect a sample at the nominal timepoint then simply recording the actual time is sufficient without having to worry about protocol deviations. This will focus sites on getting the actual time recorded in the CRF and help clinical research monitors by reducing the monitoring of time windows around PK sample collection.
Granted, it is preferable to have samples collected as close to the nominal timepoint as possible but including sampling windows doesn’t guarantee that that will occur. Overall, PK sampling windows do nothing but cause unnecessary work by the site and Sponsor.