Forum Replies Created
-
Posts
-
-
I think the CRF is very well-made! The form is easy to read. The boxes and codes are consistently used to make it easier for staff to enter the data. It also covers all necessary data without redundancy and does not ask for identifiable data.
I like the staff’s signature part, so we’ll know who recorded the data. I also like the ‘other’ choice for the sex data. I didn’t consider these things when I made my CRF, but I have learned now.
A few improvements I can think of are listed below:
– There should be a pregnancy test result from the laboratory.
– The vaccination arm should be the type of vaccine (QIV or TIV).
– The antibody titer should be recorded as a value (such as 1:______).
– The date pattern (DD-MM-YYYY) is missing from some of the date data. The unit for systolic/diastolic blood pressure should be specified (such as mmHg).
– This one is just my personal preference, but I think it would be nice if all the physical examination records were on the same page.Thank you!
-
-
-
-
I have entered data into databases a few times. The programs were Microsoft Excel and the project’s website. As far as I know, there was no edit check or logical check, but the website was well-designed, making it easier to enter data. The website also used user authentication and recorded timestamps and user names of people who entered the data.
For data backup, the project I have participated in keeps copies of data in several places, such as the institute’s server and external hard disks. Also, when we do data analysis, we try not to work directly on the original files. We always copy the files and work on the copies instead to make sure that the original files are safe.
-
I’ve done data analysis and a bit of data entry and quality control for genotype data.
If I could, I wanted to reduce the number of steps that rely on manual data entry. There was a step where staff needed to input sample IDs into the computer manually, and although we tried to recheck them thoroughly, errors might still happen. The process was also laborious and inconvenient.
Another thing is to keep records, both on paper and computer, more systematically. With several workers, it could be hard to find things, such as documents, as each person organized things differently.
-
Hello, I am Paravee from Translational Medicine Program. It is my first time doing a topic discussion here, so please tell me if I do anything wrong. I have never experienced collecting clinical data myself, but I analyzed existing data in my project. Maybe I can talk about them instead?
1. In the primary project, the clinical and genotype data were collected for research purposes. My project is to study the carrier frequency of a rare disease, so I mainly work with the genotype data.
2.For me, it was a secondary data collection because the datasets have already been collected in the previous project.
3.In the primary project, the data were collected using several methods, such as medical records, laboratory records, and self-administered paper-based questionnaires.
4.The project included many participants whose data were collected during several visits, so there are some inconsistencies in keywords/codes used across the datasets, making it complicated to combine and analyze data. Also, several steps relied on manual data entry, which increases the chance of errors in the database.
-
-
-
Thanks for sharing your answer. It is a good point, and you explained it so well! I agree that it is crucial to make the dataset understandable for everyone. There was this one time when I received a data table and couldn’t understand the column names because they were all abbreviations. Turned out there was another file that should be sent together. It was a table that explains what the abbreviations mean and what are there measuring units. So it is very important to make everyone understand the data the same way.
-
-
Thank you for sharing! Google form is a popular platform for collecting data these days, probably because it is easy to create and use. Collecting data from the form seems to be a very tough work though, so we really have to design the questionnaire well. Also, I totally agree that using an existing database can be really difficult. The data we want to use might be invalid or missing, and we can’t really do anything about it.
-
-