In life sciences, the precision of a measurement is crucial for reporting and evaluation of compliance to specification.
In the paper world of batch record entry, a number of significant figures is typically depicted for the performer to record a value to the required precision;
Take a pH sample and enter the value: (##.## pH) for example.
In many MES system EBRs, it is possible and desired to enforce the format of the entered value to comply with the record.
Performer enters 7.10532 and the system rejects the input. Many organizations instruct the performer to round and they enter 7.11 to comply. Some might direct the performer to truncate (based on the fact that the instrument is only certified to be precise to the hundredth, and the 0.00532 cannot be trusted) and they enter 7.10 to comply. Which to employ is a whole other discussion.
Rather than doing this, would it be appropriate to accept the performer's raw entry, 7.10532, then have the MES do the rounding to 7.11 or truncation to 7.10, and ask the performer to confirm the calculated value?
The rationale is that rather than force the user to make a mathematical decision that they could do in error, have them enter the value exactly as they witness it from the instrument, then let the EBR define the significant value.
Provided the performer's original entry is recorded, and they get the opportunity to confirm the rounded/truncated value, from a data integrity standpoint this should be superior to validating and enforcing an entry that is manipulated by the performer.
In reply to Keith Mattox:
In reply to Gary (Loves This Stuff) Macri: