Dose Decisions, Modifications and Errors

Part of Principal Investigator (PI) Training: Investigational Product Decisions.

About 4 minutes. Free, in your language, no account needed.

Why this matters

Prescribing inside a protocol is the strangest thing a clinician does in a trial. Your training says titrate to the patient. The protocol says follow the rule. Both are trying to protect the same person, and when they disagree you have to know which governs and why.

The rule, and its exception

Section 2.10.6 requires the investigational product to be used in accordance with the approved protocol. Section 2.5.4 says to follow the protocol and deviate only where necessary to eliminate an immediate hazard to trial participants. Together those mean: the protocol's dosing rules are not clinical suggestions.

The exception is the one from Module 8. Section 2.5.4 permits deviation without prior approval to eliminate an immediate hazard to a participant. Holding or stopping a dose because the person in front of you is deteriorating falls inside it. What does not fall inside it is the reverse: increasing a dose, extending treatment, or re-starting after a protocol-mandated stop because you think it will help.

The asymmetry is deliberate and it is worth internalising. You may always give less. You may not unilaterally give more.

Working the dose-modification rules before you need them

Every protocol with a dose-modification scheme has one page that decides most of your decisions, and it is almost never read until the first event. Find it at start-up and answer four questions in writing:

• What triggers a hold, a reduction, and a permanent discontinuation?
• Who decides — is any of it delegated to the sponsor's medical monitor, or is it all yours?
• What is the window for re-starting, and what has to be true first?
• What happens if a participant misses doses — is there a catch-up rule, or is the visit simply recorded?

Then put the answers where the on-call physician can find them at midnight. A dose-modification scheme that lives only in a 180-page PDF is a scheme nobody will follow correctly under pressure.

When the protocol and your judgment disagree

This will happen. A participant is tolerating the drug poorly but not to the protocol's threshold, or meets a discontinuation criterion but is clearly benefiting.

The sequence that works:

1. Act on safety first. If continuing poses immediate danger, hold, and document why (§2.5.4).
2. Call the medical monitor. Not for permission to protect the participant — for their read on the protocol and on what other sites have seen. This is what they are for and physicians under-use them.
3. Record your reasoning, not just your action. What you observed, what the protocol said, what you weighed, what you decided.
4. Ask whether the rule is wrong. If the threshold is repeatedly out of step with what clinicians are seeing, that is amendment evidence, and it helps every site.

What does not work is quietly managing the patient your way and reporting a compliant visit. That is a deviation plus a data integrity problem, and it is the pattern that ends careers.

Medication errors

Wrong dose, wrong participant, wrong product, expired product, missed dose, dose given after a discontinuation. Treat every one the same way and it stops being frightening:

• Attend to the participant. Clinical management first, always.
• Report it promptly, per the protocol. Most protocols require notification whether or not harm occurred, and an error with no harm is still an error.
• Assess whether an adverse event occurred, and record it if so. The error and any resulting event are separate records.
• Consider unblinding only on the Module 8 test — would knowing the allocation change this participant's management now? An overdose requiring an arm-specific intervention usually meets it.
• Document the same day, and find the mechanism. Errors cluster around handovers, look-alike packaging, and busy clinics.

The instinct to minimize an error where nobody was harmed is strong, and it is the reason a second one happens.

Key takeaways

• §2.10.6 plus §2.5.4: protocol dosing rules bind you. The asymmetry to remember is that you may always give less on safety grounds, never unilaterally more.
• Work the dose-modification page at start-up and put the answers where the on-call physician can reach them.
• When protocol and judgment conflict: act on safety, call the medical monitor, record the reasoning, and feed a repeatedly wrong threshold back as amendment evidence.
• Report every medication error including harmless ones, record any resulting adverse event separately, and fix the mechanism.