Whenever possible, trials should use a computerised system for
data
management.
Such computerised systems are useful to keep track of the flow of
participants in the trials (e.g. recruitment, withdrawals, reason for
withdrawals, data collection, and for recording the flow of information
to and from participants (e.g. questionnaire dispatch and
return)). These
systems also enable quick and efficient data validation and quality
control and may also be used to set up standardised management reports
for the Trial Management Groups (see Trial
management).
An effective data management system needs to be flexible and adaptable,
so that it can be tailored towards the needs of the people
collaborating in
the trial. For larger trials this may be a customised
programme,
but
smaller trials may adapt off-the-shelf software. The most
suitable data management system will depend on budget constraints, the
complexity
of the trial and the technical environment. Asking the advice
of
an
experienced trial programmer can be invaluable in helping to choose the
right
system.
Epi Info is a public domain software package designed for the global community of public health practitioners and researchers. It provides for easy form and database construction, data entry, and analysis with epidemiologic statistics, maps, and graphs. The primary applications within Epi Info are:
This figure was contributed by Eduardo Bergel who prepared it for the Perinatal Care Trial.
Data collection forms (e.g. CLIN1, CLIN2, ALTA1 and ALTA2) are modified perinatal clinical history forms (PIS) that incorporate data particular to the study and hide the rest of the data. These forms are generated as a carbon copy of the PIS. Every morning, the data clerk at hospital level collects all CLIN forms for deliveries fo the previous day and all ALTA forms for all women that were discharged from the hospital that day. The data manager enters the data in the Remote Data Management Subsystem (RDMS) for that hospital. The RDMS is a stand alone application that runs on a personal computer. Consistency checks are performed during the data entry and warnings are displayed. The data is checked to make certain that entered values are acceptable, that all required fields are entered, and that items are consitent with other related items in the database.
Once data editing and any error resolution is completed data and validation reports are sent using XML messaging system from hospitals to the coordinating centre, the Latin American Centre for Perinatology (CLAP) on a daily basis. At CLAP, the data is reviewed, validated, stored in back up files. Validated databases are then sent to the Research Triangle Institute (RTI), where the main study database is held.
|
Eduardo Bergel, WHO, Switzerland, explains the different approaches trialists can take to data management (MP4, 2 minutes 18 seconds). Can't see the audio controller? Play it in your media player. |
What options are there for data management systems? Eduardo talks about the open source system OpenClinica (MP4, 5 minutes 49 seconds) Can't see the audio controller? Play it in your media player. |
What are the key things to think about with regard to data management? (MP4, 1 minute 54 seconds). Can't see the audio controller? Play it in your media player. |