Learnaboutgmp Community

Environmental Monitoring issue

We are starting to put down some initial User Requirements for a
environmental monitoring software system and have a few questions on
entering result data directly into the system without first writing
down the results on paper and signing it.

Can the results from a lab analysis (For example: a visual count of
colony formation) be directly input into a computer system by the
person performing the count without first being written down on
paper. If so.

  1. Not having a physical piece of signed paper anymore, would the
    input be considered as the original “raw data”.
  2. To eliminate typing errors does the result have to be re-entered
    into the system by a second operator or would a review of the data
    entered by a second person suffice.
    3.Not having a signed piece of paper anymore would each electronic
    record need to be electronically signed (electronic signature) or can
    it be regarded just as an electronic record.

Basically we are starting to put down some user requirements and the
idea is to have two qualified lab technicians verify the results and
enter them directly into the computer system. We are playing around
with the idea of either:
1 Double data input: i.e… The computer system would compare the data
entered by technicain 1 and technician 2 and highlight any
discrepancies. (this should eliminate typing errors and also confirm
the result).
2. Single data input. The second technician would review the data and
confirm the result by entering username and password.

Any comments on how you manage this sort of data input would be
welcome.

Directly entering data into the software with appropriate procedural instructions regarding the verification is the way to go. Having a second person check the typed data defeats the purpose. Writing it on a piece of paper first and then entering the data will cause more problems. Additionally, if your software is able, put some decision rules into the software to force a double check for unusual results. For example if you are counting colonies and you have an alert or action limit for the counts, at those limits a flag (color change or pop up box) can alert the person entering the data to check the validity of the data. This would be an additional check to the procedural process described.

Again, you should not have to put extra steps into the process that are not in the paper world.