Medical Coding Service

QA Data provides a manual Medical Coding service that ensures all medical terms, drug references and indications are correctly defined and categorised to allow accurate analysis and review. This is an important task that requires human judgement, experience and precision.

What is Medical Coding?
Medical Coding is the process of ‘grouping’ and ‘classifying’ reported medical terms, drugs or indications, assigning each one to its correct standard medical term, classification or coding.

To be effective, Medical Coding must be:
• Accurate – coding to the closest possible term
• Consistent – similar reported terms all coded the same way

QA Data’s Medical Coding service inspects all the data collected during a clinical trial and ensures they are correctly described, labelled and categorised. This is important for the data to be collectively analysed and reported and to provide standardised reporting for regulators.

Our service is based on the definitive medical dictionaries: MedDRA and WHOART for adverse event (AE) and disease terms, with all drug terms being aligned with the WHODDE dictionary. We code to the closest possible term, making sure that similar reported terms are all accurately identified and consistently grouped and classified.

Challenges
When clinical trials are conducted across multiple sites with different investigators, it is a real challenge to ensure uniformity. The task of the statistician in counting and reporting on AEs, disease indications or drug terms can seem impossible when different descriptions or terms are intermingled. Left unaddressed, it can also make it almost impossible for regulators to review the data with any confidence.

An example would be where one site reports a patient with ‘headache’ whereas a second describes the same indication as ‘sore head’. Similar confusion can occur between such terms as ‘hypertension’ and ‘high blood pressure’. In fact, these describe identical events but failure to count them as such could have significant implications for the sponsor as well as for patient safety.

An even more worrying example would be for different sites reporting ‘Endometrial Stromal Tumor’ and ‘Endometrial Stromal Tumour’, with both classified as ‘Endometrial Stromal Sarcoma’. This would be consistent but far from accurate since sarcoma is a very specific type of tumour. Both terms should be reclassified as ‘Endometrial Neoplasm NOS’ since it is safer to select a less specific but correct term.

Why not Auto Code?
It is theoretically possible for all coding to be performed by automated programs. Auto coding relies on loading clinical trial data into the program, which then seeks a matching lower level term (LLT). However, performing reliable auto coding requires the CRF data to be manually ‘cleaned’ eliminating all errors in spelling, punctuation, spacing, upper or lower case, before matching can begin. And even then, as you can see from the examples above, correct coding can depend on human judgement, knowledge and experience.

There have been issues with auto coding results needing to be recoded, as the software has been unable to differentiate between subtle yet significant medical differences. There are also many nuances and issues with electronic auto coding systems known to ‘under code ’ (leave part out) or ‘over code’ (attach to higher level event).

An example of under coding would be to assign ‘Coronary Atherosclerosis with Recurrent Angina Pectoris’ to the LLT of ‘Coronary Atherosclerosis’, rather than requesting the site to split the events. Over coding is exemplified by assigning ‘Pneumonia Left Lobe’ to the LLT of ‘Lobar Pneumonia’ under the erroneous belief that ‘lobe’ and ‘lobar’ are synonymous.

The Rigorous Approach
While auto coding and inexpert manual coding continue prone to such problems, QA Data will continue to offer its expert Medical Coding service:

• Critically analysing and interpreting all reported terms
• Accurately making a precise and correct selection from the myriad of possible choices
• Consistently assigning the term to the highly specific dictionaries used today

In doing so, we will greatly enhance the reporting from our clients’ clinical trials and their chances of being fairly and accurately reviewed by regulators.

To learn more about our Medical Coding Service, please contact QA Data directly.