Exactly Means Exactly

Part of Research Assistant (ASSIST) Training: Following an Instruction Exactly.

About 2 minutes. Free, in your language, no account needed.

The protocol is not a guideline

A protocol says things like: take the sample between 30 and 45 minutes after the dose. Record blood pressure after the participant has been seated for five minutes. Use the second reading.

These are not suggestions or rough targets. They are the conditions under which the result means something. A sample taken at 50 minutes is not a slightly worse version of the right sample. It is a different measurement.

Why small changes matter more than they look

Drug levels in blood change by the minute. Blood pressure after two minutes seated is genuinely different from blood pressure after five. If some sites wait five minutes and some wait two, the study is comparing numbers that are not comparable, and no statistician can repair that afterwards.

The value of a trial comes from everyone doing the same thing the same way. Your consistency is part of the result.

Read it again, even when you know it

Experienced staff re-read the instruction for a task they have done fifty times. Not because they have forgotten, but because protocols get amended, and the version you learned may not be the version in force.

Check you are working from the current version. If you are unsure which that is, ask before you start rather than after.

If you cannot do it as written

Sometimes you cannot. The participant arrived late, the equipment is broken, the window has passed.

The answer is never to do it approximately and say nothing. Tell the coordinator before you act if there is time, and immediately afterwards if there is not. What happened gets written down as it happened. A recorded departure from the protocol is manageable. An unrecorded one is not.