Medical Coding in Clinical Data Management

In any clinical trial conduct, recording and storing data in a controlled, consistent, and reproducible manner for data retrieval and analysis is a necessity for regulatory compliance and clinical study success.

What is Medical Coding?

Medical coding is the classification of multiple similar verbatim terms, using a validated medical (or medication-based) dictionary supplied by the customer, or under license by the relevant licensing bodies (MSSO, Uppsala), in order to produce a statistically quantifiable count of all similar terms in a given database.

Along with data entry, validation, data processing, reconciliation, external data load, and many more clinical data management related activities performed in Clinical Data Management systems (CDMS), medical coding is performed to facilitate the summarising and analysis of certain sets of data (e.g. Adverse Events, Medical History records, Concomitant Medications etc.). To provide control and consistency, a variety of medical coding dictionaries may be used to process, analyse, and report collected data. The coded variables/terms are used by sponsors/medical monitors to review the events and medications throughout the study as appropriate.

Study statisticians and medical writing groups use the coding reports to get the quantitative numbers which is included in the corresponding sections of the TLFs (Tables, Listings & Figures) generated for the study which is eventually reflected in the Clinical Study Report (CSR) created for regulatory submission.

With multiple versions of medical dictionaries released by the managing bodies every year, processes must be established for managing the release of multiple versions of the same dictionary, handling different dictionaries or versions that have been used, and integrating data coded with different dictionaries or versions.

Standard Dictionaries Used for Medical Coding

Two of the most commonly used dictionaries are:

Medical Dictionary for Regulatory Activities (MedDRA)

MedDRA: Coding of Adverse Events and Medical History events using this dictionary is required to group data for meaningful analysis. MedDRA is the ICH-developed and recommended dictionary for all medical events captured in clinical trials, including, but not limited to, AEs and medical history terms etc. MedDRA has multi-axial functionality and provides multiple levels of terms and codes which require a distinct understanding by the coder to pick the correct code. Coders and reviewers of medical information must have an understanding of the flexibility of MedDRA as well the implications that its storage and implementation can have on safety reporting.

The levels of terms used in MedDRA are as follows:

• Lowest Level Term (LLT)
• Preferred Term (PT)
• High Level Term (HLT)
• High Level Group Term (HLGT)
• System Organ Class (SOC)

MSSO (Maintenance and Support Services Organisation) is the organisation responsible for publishing and maintaining MedDRA. MSSO releases two versions annually:

• 1^st March X.0 release (changes in LLT, PT, HLT, HLGT and SOC).
• 1^st September X.1 release (changes in LLT and PT only).

The upgradation mainly covers retirement of terms, addition of new terms identified and approved, and updating of the assignments to SOC and consistency of available terminology. In MedDRA, a preferred term (PT) may be associated with multiple SOCs. However, each PT is associated with only one primary SOC.

The latest available version of MedDRA is 27.1 which was available from 1st September 2024. Coding specialists can be certified via the MSSO Certified MedDRA Coder (CMC) exam.

World Health Organization Drug Dictionary (WHO Drug or WHODD)

WHO Drug Dictionary works in a similar way to MedDRA but is used to code medications. Verbatim terms (the medication names) are coded to a hierarchy of terms (Preferred name and to ATC level 4), as demonstrated below in our table.

ATC assignment requires the indication and/or dose and/or route to be available. If insufficient information is available to code the ATC accurately, the coding specialist needs to get additional information from site to select the correct code. As per standard practice, if additional information is not available and an ATC classification is required, the most common ATC classification is assumed. This should be agreed upon by the sponsor beforehand and documented.

If high level ATC classification (level 4) is to be performed for a project, any term assigned a multi-ATC code needs to be manually assigned by the coding specialist using the dose, route, and indication of the associated drug. Terms will be coded to the highest level of specificity possible.

World Health Organization (WHO) designed the WHO Drug Dictionary for medication coding. In 2005, the Uppsala Monitoring Centre (UMC) introduced the WHO Drug Dictionary Enhanced (WHODDE) Browser. WHO-DDE combines data from the original WHO Drug Dictionary (WHO-DD) with additional country-specific drug information. UMC is responsible for maintenance and publishing the dictionary. UMC releases the WHO Drug Dictionary twice a year i.e. March and September which can be used by organisations as per subscription. The next WHODrug Dictionary release is due in March 2025.

In September 2020, final version of WHO Drug Enhanced was released and phased out in favor of the standardised and more comprehensive WHO Drug Global dictionary. From March 2021, the only dictionary available from UMC is WHO Drug Global.

WHODrug is distributed to all users in two formats – the B3-format and the C3- format.

The B3-format contains information about trade name, ingredients and ATC classification(s). The C3-format contains all the B3-format information (including the Drug Code). In addition, it has information regarding the countries in which the product is marketed, Marketing Authorisation Holders, pharmaceutical forms and strengths. Other dictionaries available for medical coding are:

• COSTART (Coding Symbols for a Thesaurus of Adverse Reaction Terms)
• ICD (International Classification of Diseases)
• WHO ART (WHO Adverse Reactions Terminology)

Medical Coding Tools and Methods

Medical coding is performed using the dictionaries installed in the software applications. Coding specialists work on this tool to assign the appropriate codes to the terms. The features of the tool or the standard processes per which coding activity occurs are as below.

1. Auto encoders: A programmatically assisted process for matching a reported term to a dictionary term. In this process, the tool runs the validation during which the verbatim terms with exact match present in the dictionary gets auto-coded.
2. Manual Coding: In the manual coding process, the coding specialist selects an appropriate dictionary entry for each reported term in the coding tool. The coding specialist should be able to raise queries and use the search feature of the tool for efficient coding. The coding tool or applications should be validated systems as per regulatory requirements with audit trails in place. In manual coding processes, the coding reviewer uses the tool to review the coded terms for accuracy and consistency.
3. Hybrid Approaches to Coding: This is the most standard approach followed in coding set up in which the reported verbatim terms are first automatically coded to the exact match or that match a term that has previously been coded (i.e., added in synonym list) present in dictionary using an auto-encoder. The terms that are not auto-encoded are then manually coded by the coding specialist.

Artificial Intelligence (AI) in Medical Coding

AI in medical coding has made significant advancements, streamlining and improving the accuracy of coding processes in clinical research.

1. Automation and Efficiency: AI-powered tools are being utilised to automate the medical coding process, reducing time and effort required by coders. These tools scan clinical documentation and automatically assign appropriate medical codes to the verbatim terms.
   
   
2. Improved Accuracy: AI programs, especially those using natural language processing (NLP) and machine learning, have improved the accuracy of medical coding by understanding complex medical terminology and context. This reduces the risk of human error and enhances compliance with regulatory standards.
   
   
3. Insights and Data Quality in Medical Coding: AI not only assists with coding but can also improve the data quality by identifying inconsistencies, gaps, and incomplete information within the medical records. This leads to more accurate claims processing and better data for clinical and financial reporting.
   
   
4. Real-time Coding: Some AI tools offer real-time coding suggestions, allowing coders and clinicians to address discrepancies immediately, further reducing delays in claim submissions.
   
   
5. Challenges: While AI has great potential, challenges remain in achieving 100% accuracy, handling ambiguous cases. There are also regulatory concerns and the need for ongoing oversight by users to ensure quality.

Overall, AI in medical coding management is enhancing efficiency, accuracy, and compliance, though human expertise is still crucial for quality assurance and complex cases.

Reporting of Terms

For efficient and correct coding, it should be ensured that the verbatim terms recorded by site are specific. The conventions should be agreed upon by the site beforehand and documented. Some of the examples are listed below:

• AE term reported should not be ambiguous i.e., the event recorded should have a clear meaning.
• Two or more terms should not be reported as one verbatim, e.g. ‘diarrhoea and vomiting’, these may be indicative of one diagnosis.
• Symbols and abbreviations should be avoided while recording AE terms as these may get misinterpreted.

After the recent MedDRA & WHODDrug user group conference held in Bangalore on 6th February 2017 there are several points to consider from a coding perspective:

• WHODD has been re-named as WHODrug products (re-branding)
• If WHODDE is mentioned in the file it means it is an enhanced version, if codes are present within the PT term it means B2 format.
• The first release of full B3/C3 format is available from 1st March 2017.
• In B3/C3 format, the truncation issue is resolved which has been observed in B2/C format, the field length has been increased to 1500.
• Validation process for each dictionary versions to be done for clients, the sponsors should also have WHODrug versions which the CRO is working on - this is a mandatory step to be followed for dictionary validation.
• In B3/C3 format, the preferred names are generic.
• MSSO provides free training webinars, website to be checked for details.
• As a licensed subscriber, 100 change requests can be sent to MSSO per month, however they will be approved/rejected based on MSSO’s judgment.
• Self Service Application (SSA) was released in April 2017.

Partner with Quanticate for Expert Medical Coding and Clinical Data Management

At Quanticate, we recognise the vital importance of precise medical coding and efficient clinical data management in the success of your clinical trials. As a leading global Biometric Contract Research Organisation (CRO), we specialise in transforming complex clinical data into meaningful insights using industry-standard dictionaries like MedDRA and WHO Drug Global. Our team of experienced coding specialists ensures that your data is coded accurately and consistently, facilitating effective analysis and reporting.

Leveraging advanced technologies, including Artificial Intelligence (AI) and Machine Learning (ML), we enhance the accuracy and efficiency of the coding process. Our expertise extends across various therapeutic areas, ensuring that we stay current with the latest dictionary updates and regulatory requirements to provide you with the highest quality services.

In addition to medical coding, Quanticate offers a comprehensive suite of data management services—ranging from data entry and validation to reconciliation and external data integration. Our holistic approach ensures that every aspect of your clinical data is handled with integrity, precision, and in compliance with regulatory standards.

Ready to optimise your clinical data management and medical coding processes? Contact a Quanticate team member today!