Clinical Data Management – getting the basics right

Hello all,

In my endeavor to connect the technology advancements to how the clinical trials would be conducted in future, I realized that it is often the purpose of making a process simple or easier is a key driver e.g. using “bring your own device” concept – BYOD is aimed at simplifying the conduct of clinical trials in a global environment. There are many such examples that are being adopted now. However, looking at such instances made me realize that even for adopting the new technologies it is important to get the basics right.

data collectionThe basics of data collection design and data cleaning still stand tall and are fundamental. The fact that sites get confused with the complexity of data entry and when attempting to answer some data cleaning queries; itself indicates there is huge scope of improvement.

In my opinion we should think “COMPACT”. It is the buzz word for getting the basics right.


C = Common Sense –
Common sense should prevail while designing a data collection tool commonly called as Case Report forms (CRF). Lot of thought should be put together not only to understand the need of collecting data but alsoCommon Sense to eliminate unnecessary data. A guide to arriving at this decision is to have an understating of the “must have”/required data collection items/variables as per the contemporary data standards. An example could be of the SDTM model and SDTMIG which indicate items/variables that are required.

Obvious
O = Overcoming the Obvious trap
– How many times as data manager we end up thinking isn’t it obvious? Why was the data entered in the wrong way? Why is the query not getting answered in first instance? Well, over the years I have understood that what is obvious for a data manager may not be, in fact it is not so obvious to the clinical trial site personnel. It is the intuition of what the clinical trial site personnel would record in the CRF that will help make a perfect data collection tool. An example is use of drop down lists. While these lists facilitate consistency in data collection, it is important to ensure that useful clinical information is not lost in the process.

MeasureM = Measure – To be compact and avoid the obvious trap, we need to measure at a TA level the data collection items/variables which led to inappropriate data entry. These data collection items/variables must be scrutinized further and the analysis presented on what type of data entry errors have been observed. The errors can be broadly grouped into Attribute errors, Layout errors, and Concept errors. The commonly seen attribute errors stem from length restriction, precision restriction, date restrictions etc. The length and date restrictions are an annoyance to the site. The current methods of avoiding these data entry errors such using additional drop down list to accommodate unknown dates and data are not intuitive enough to prevent the data inconsistencies.The layout errors are easy to fix and if possible a data collection page should not run into multiple subpages. The risk of missed data entry and erroneous data increase if data collection module runs into multiple subpages. The concept errors are most difficult errors since these are often scientific and the purpose of research. These errors usually surface when the site attempt to accommodate information in the available data items due to lack of space to fill this information. It may not be a best practice to accommodate this information as PI comments. If one carefully analyzes the PI comments, these comments would indicate need of additional data collection items.

PurposefulP = Purposeful – The purpose of designing a data collection tool is driven by the science captured in the clinical trial protocol. To create a purposeful data collection module, one must understand the science in the protocol. It is most important to understand the Primary and Secondary endpoints, the assessments  in the clinical trial protocol. Putting this understanding to use to collect data according the clinical trial protocol will help create a purposeful data collection tool.

ACT – Action – The influence of technology is inevitable, it is important to act now to consciously develop the expertise to do the basics right. These are fundamentals which are less likely to change. The art of writing accurate specifications must be developed and one should invest in developing this skill. Any technology would require best in class specifications that help creating a lucid data collection tool

There will be more on the data cleaning in the next post…until then…think basic…think right…

@disrutpcdm

Advertisements

I am a clinical data management professional with 13 years of experience in healthcare and clinical trial data management. I am focused on bringing disruption in the area of clinical trials by conceptualising break through data management practices.

Posted in Best Practices
4 comments on “Clinical Data Management – getting the basics right
  1. prachi shah says:

    Great writing Dr. Abhishek, very basic yet very important. Thank you for sharing.

    Like

  2. […] clinical data management scope of work, we already discussed data collection basics; the next most important work is to validate the collected […]

    Like

  3. […] manager must take a bigger role in creating an error free new standard by following the basics of CRF Creation and Data validation. This will ensure no […]

    Like

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Enter your email address to follow this blog and receive notifications of new posts by email.

Archives
%d bloggers like this: