...
How do I know how many ‘Item Group Data’ I will use?
A: There is no definite answer to give to this question, as it will depend on many factors that can change between units. But it is possible to make an estimate. To do so, we advise to multiply the expected number of assessments with a scaling factor of 1.4. This scaling factor is an empirical number and may not be accurate in all situations, but gets a realistic estimate.
What is a scaling factor and why does it need to be applied?
A: In real-life we see that units collect more data than just the bare assessments. For example, a protocol may dictate that volunteers lie down 10 mins before their ECG is taken. If the unit decides to record that event as source data, it would trigger an Item Group Data in addition to the actual recording of the ECG. So while many would consider this to be one assessment, in ClinSpark two Item Group Data are recorded. Also, some data are collected as one assessment, but because of CDISC data structure requirements, this would be recorded as multiple Item Group Data. Medical History is an example of that. So when enrolling a subject into a study who is new in the database, recording the Medical History may result in multiple Item Group Data. However, once the subject is in the database and its data is pulled into the next study, no new Item Group Data are generated. All these effects together make the number of Item Group Data typically a bit higher than the number of assessments, something that we capture in the scaling factor.
...