groups » Data Management and Statistics » Ask The Expert - Data Management Month

During March 2013, we are very fortunate to have Lesley Workman of the University of Cape Town on hand to answer any queries you have about data management, in addition to our data management expert committee.

Lesley is currently employed as a Scientific Co-ordinator and Data Manager for the Clinical Pharmacology Module of WWARN (World Wide Antimalarial Resistance Network), the main focus of her work is in standardising individual patient anti-malarial PK-PD data contributed to WWARN for inclusion in pooled analyses. She completed her MPH at UCT in 2009 (Epidemiology and Biostatistics track) while working as Data Manager for the South African TB Vaccine Initiative (SATVI).

She has been involved in clinical research since 1981, the last eleven years being in data and co-ordination management of clinical research studies in TB, HIV and Malaria where she has developed and managed large databases for a number of projects using MS Access and commercial Clinical Data Management Systems.
In addition she has completed an advanced Data management course which focused on the regulatory requirements for early phase studies and has completed the user training for OpenClinica.

Her overall vision is to work towards a standard Clinical Data Management System suitable for an Academic Institute that will comply with all the necessary regulations that apply to registration of a new compound with the FDA.

Lesley is on hand to answer ANY questions you have about data management - which databases to use, what is involved in data cleaning, eCRFs versus paper, how to progress your career, what training is available - anything you like. Please feel free to post any questions below!

Reply

data entry data management

  • marietjie_pretorius_uct_ac_za Marietjie Pretorius 12 Mar 2013

    Lesley, we experience problems with the management of our data during the study. For e.g. the screening log, the data that we need to see how far we are. We want to compile a workbook and then this morning I see this email from GlobalHealth. Your overall vision, is also our vision here at the LungInfection Unit, UCT. How can we proceed? What is the possibility to go and see you? Kind regards

  • gertrudemutonyi MUTONYI GERTRUDE 12 Mar 2013

    Lesley, I am happy to learn that you are on board to answer any data management questions. I would like to learn more about locking my database so as to improve on data security. I mainly work with MS Access databases though there is a plan for us to train in OpenClinca as well. I also have a question on how to handle time in MS Access. I designed a table with time variables and I set it to short time and included and input musk. However when the data was transfered to STATA it didn't seem that meaningly as it did in Access. Please advice, regards
    Gertrude

  • mutebe2 ALEX MUTEBE 12 Mar 2013

    Hi Lesley,
    What is the procedure for creating a data management SOP. Thanks

  • lbufumbo Leonard Bufumbo 12 Mar 2013

    Dear Lesley,

    I have used Personal Digital Assistants - PDAs for data collection. We used pendragon software to sync data onto computers. If am to use a smartphone because it is more recent technology, what software do i have to use to sync data onto the computer? pros and cons? which is the easier and better software?

  • @ marietjie
    It will easiest to set up a meeting to discuss this, I suggest importing your screening log into MS Access and then using some queries to report as needed. Do you have a free hour this afternoon/Friday morning?

  • @Mutonyi
    1. Data base lock, I will post a link to an SOP on database lock with appropriate logs on the GHT website later today and send you a note when it is up.
    2. Good that you have set your time to short time in MS Access, I find it best to do all my time calculations in MS Access before importing to STATA (often converts date/time variables to a string, but we can format the variable in STATA, it will need to a double and stored as a %tc). Let me know if you would like some guidance on these time calculations
    3. To reformat a string variable to time variable in STATA:
    a. Right click on the time variable EG “timeofbirth” (in the variables window)
    b. Select Format “timeofbirth”
    c. Type of data select clock

  • @Alex
    GHT has a number of very useful documents here:
    http://globalhealthtrials.tghn.org/articles/data-management/
    For more specific SOPs on DM this link will take you to the WWARN site where you can download template SOPs and logs, these will need to be tailored to suit your system/specific study requirements
    http://www.wwarn.org/toolkit/procedures/pharmacology
    1. Database construction and security
    2. Data receipt, entry and validation

  • @Leonard
    Hi I have referred this post to Mike who has a experience in these kind of technical issues

  • @Leonard
    I have not used Pendragon, but here goes.
    Option 1) Keep your existing infrastructure and investment in Pendragon.
    According to Pendragon website, the latest version Pendragon VI will run on Apple iPhones and iPads and Android phones and tablets such as Samsung Galaxy. Palm PDAs use Pendragon 5.1 and there are some issues with using Pendragon 5.1 forms in Pendragon VI. These are documented at http://pendragonsoftware.com/formsviinfo/faq.html.

    Option 2) Throw away Pendragon and go with the latest gee-whiz technology. There are so many options out there, some specific to Apple, some to Android, and all doing things differently.

    So unless you don't like Pendragon and want to change to something else, I suggest you take option 1 and upgrade to Pendragon VI. You can get a 14 day free trail, so maybe that's a good starting point to see what works without having to change anything. Then get in touch with your Pendragon agent and start asking the difficult questions about keeping your current database and server and using the new Pendragon VI with it.

    Unfortunately that's the best I can do remotely. Contact me if you'd like further support.

    Regards

    Mike

  • lbufumbo Leonard Bufumbo 13 Mar 2013

    Thanks Mike,

    Regarding option 2, are there any free trials that I can run before investing? Lets say for an Android phone?

    Thanks again
    Leonard

  • Hi Leonard
    There is just so much out there in the way of Apps for Android, many of them general purpose apps, some very specific. I'd need to know a lot more of your use of Pendragon (what's its purpose, what data do you record, what do use the data for afterwards, how many people and what skill levels. I'll send you a connection request and we can take this off-line to avoid overwhelming the others here.
    Regards
    Mike

  • @Mutonyi
    The SOP for database lock is included with this template SOP (PHA_001) http://www.wwarn.org/toolkit/procedures/pharmacology

    You will find the example form 5 here as well. Sometimes a "soft lock" is used, this is good practise. Once you are happy that all outstanding data queries have been resolved, edits checks run and the data entry has been completed, soft lock the database(without removing access to the data), run the analysis tables so that if data queries come up during the analysis, we can respond to these without having to formally unlock the DB.

  • mbirungi Molly Birungi 21 Mar 2013

    Thank you for this mail and the opportunity to ask some questions.

    I want to help me with the following
    • How do I delete a record permanently from the data base, other than replacing it or marking it for deletion.
    I am using EPi Info access based.
    • Which is the fastest way of dealing with multiple entries of the same record (other than replacing)
    • I also need to help me with information on how to analyze my records/cases received in a particular week. I want to analyze (tracking) for example number of cases received in a particular week. or in the various weeks of the month.

    Eagerly waiting for a response
    Thank you

    Molly
    EPI laboratory, UVRI

  • @Molly
    1. The only way to permanently delete a record in your database is to identifiy the record in the relevant table and delete the entire row.
    2. To manage duplicates, the best way is to delete the duplicate record as above once you are sure this is a real duplicate. Another approach is to mark as a duplicate and when exporting data for reporting/ analysis or managemet to not include these.
    3. For the reporting on progress during a study, this will be best done by developing some standard outputs - I will investigate and get back to you with some suggestions.

  • GHN_Editors The Editorial Team 28 Mar 2013

    Thank you to everyone who has participated in the 'ask the expert panel', and especially to our Lesley and Mike for their expertise.

    Following the very successful discussions here, the Global Health Network has launched a specific area just for discussing data management. Led by a team of specialists, the area has recently been launched and you are invited to use it to share resources and templates, network with one another, make suggestions about what the site can offer you, and discuss events and ideas.

    The site is called the Association of Data Management in the Tropics (ADMIT) and you can find it here: http://admit.tghn.org/. You can also access it, as with all the other Global Health Network sites, via the grey button on the bottom left hand side of the screen.

    As you are already a member of the Global Health Network, you should automatically be a member of the ADMIT area as well - but if you have problems accessing the site, don't hesitate to let me know at admin@globalhealthtrials.org.

    I very much hope that you will find the site useful and a valuable place to share ideas and knowledge. As with all the platforms on the Global Health Network, it will grow and change, and value will be drawn from the discussions and shared knowledge and resources made by the members - so do contribute!

Please Sign in (or Register) to view further.