MedDRA Coding in Pharmacovigilance: Principles, Process, and Common Challenges

Pharmacovigilance relies on accurate and consistent safety data. However, information in an individual case safety report rarely arrives in a standardised form. A patient may report “feeling sick,” a physician may document “nausea,” or a medical record may contain a specific diagnosis. To analyse these different descriptions consistently, pharmacovigilance teams need a standardised medical terminology.

This is where MedDRA coding in pharmacovigilance becomes important.

MedDRA, the Medical Dictionary for Regulatory Activities, provides a standardised terminology for describing medical information in clinical and post-marketing safety data. Coding allows reported information to be represented using standardised terms that can later be retrieved, analysed, compared, and presented across safety databases.

For PV teams, however, MedDRA coding is not simply a matter of finding a similar word in a dictionary. The coder needs to understand the reported medical concept, preserve the information provided by the reporter, avoid adding assumptions, and apply consistent coding conventions.

What Is MedDRA Coding in Pharmacovigilance?

MedDRA coding is the process of assigning standardised MedDRA terms to medical concepts reported in safety information.

In pharmacovigilance, coding can apply to much more than adverse reactions. Depending on the case and the database, relevant information may include adverse events or reactions, medical history, indications, investigations, procedures, medication errors, product quality issues, and other reported medical information.

The goal is to represent the information in a standardised way without changing its meaning.

For example, different reports might describe a patient’s condition as “high blood pressure,” “elevated blood pressure,” or “hypertension.” A MedDRA coder reviews the reported information and selects the term that best represents the medical concept in accordance with the applicable coding conventions.

This standardisation makes it possible to analyse safety information across cases even when the original wording differs.

The MedDRA terminology itself does not provide a complete set of instructions for every coding situation. The MedDRA Term Selection Points to Consider document provides conventions intended to promote accurate and consistent term selection, while medical judgment remains important when interpreting individual reports.

Understanding the MedDRA Hierarchy

One of the first concepts a PV professional needs to understand is the MedDRA hierarchy.

MedDRA has five hierarchical levels:

Lowest Level Term (LLT)

Preferred Term (PT)

High Level Term (HLT)

High Level Group Term (HLGT)

System Organ Class (SOC)

LLTs are the most specific level used for data entry. They represent how a medical concept can be expressed in MedDRA and are linked to a PT.

PTs represent distinct medical concepts and provide a standardised level for data analysis. Above the PT level, HLTs and HLGTs group medically related concepts, while SOCs provide broad categories.

For example, a coded medical concept may follow a structure such as:

LLT → PT → HLT → HLGT → SOC

Importantly, HLTs and HLGTs are primarily grouping levels for data retrieval and presentation rather than coding levels.

This hierarchy matters because the same coded information can later be analysed at different levels of specificity. A PV professional reviewing individual cases may need the PT, while a broader safety analysis may examine groups of related PTs at the HLT, HLGT, or SOC level.

How the MedDRA Coding Process Works

Effective coding starts with the source information, not with the MedDRA browser.

The coder should first understand what the reporter actually described. This may require reviewing the full case narrative, source document, medical history, laboratory results, and other available information.

A practical coding process can be summarised as follows.

1. Review the reported information

Start with the original wording and the surrounding context.

A single phrase may not provide enough information to identify the intended medical concept. The narrative may clarify whether a term represents a symptom, a confirmed diagnosis, a laboratory finding, or another type of information.

2. Identify the medical concept

Determine what is actually being reported.

For example, “patient experienced shortness of breath” describes a symptom. If the report instead states that the patient was diagnosed with a specific respiratory condition, the coding approach may be different.

The key principle is to code what is reported rather than what the coder believes might have happened.

3. Select the appropriate LLT

MedDRA conventions generally recommend selecting the LLT that most accurately reflects the reported verbatim information.

The selected term should preserve the meaning and level of specificity available in the source. The coder should not automatically choose a more specific diagnosis simply because it appears medically plausible.

4. Check the hierarchy

After selecting a term, reviewing its position in the hierarchy can help confirm that the selected concept is appropriate.

The hierarchy above the LLT can reveal how the term maps to its PT, HLT, HLGT, and SOC. MedDRA guidance specifically recommends checking the hierarchy when selecting terms.

5. Apply quality checks

Finally, the coding decision should be reviewed in accordance with the organisation’s procedures.

Quality checks become particularly important when several coders or teams work with the same safety database. Consistent conventions help reduce unnecessary variation between cases.

Do Not Add Information That Was Not Reported

One of the most important principles in MedDRA coding is also one of the easiest to overlook.

A coder should not add a diagnosis that was not actually reported.

Consider a report describing abdominal pain, along with elevated serum amylase and lipase. A coder should not automatically code pancreatitis simply because the findings could be consistent with that diagnosis.

The reported information should be represented using the appropriate terms rather than replaced with an assumed diagnosis. MedDRA’s coding conventions emphasise selecting terms to represent reported information while avoiding the addition of information not present in the report.

This principle is particularly important in pharmacovigilance because coding should not alter the clinical meaning of the original safety report.

The same principle applies when a report contains symptoms without a confirmed diagnosis. If the source only describes symptoms, the coder should not convert them into a more specific disease concept without supporting information.

Coding Diagnoses and Symptoms

Another common coding challenge is deciding whether to code a diagnosis, its symptoms, or both.

If a report provides a confirmed diagnosis along with characteristic signs and symptoms, MedDRA conventions may favour coding the diagnosis rather than separately coding each symptom.

However, the appropriate approach depends on the information actually reported and the applicable coding convention.

This distinction matters because coding every symptom in every case can create unnecessary duplication, while coding a diagnosis that was never reported can introduce information that does not belong in the case.

The objective is not to maximise the number of coded terms. It is to represent the safety information accurately.

Why Consistent Coding Matters

Two coders can sometimes encounter the same medical information and choose different terms. Some variation may arise from the complexity of medical language, but uncontrolled variation can affect the quality of pharmacovigilance data.

Consistent coding supports several important PV activities.

ICSR processing

Standardised terms make individual case data easier to retrieve and review. They also help organisations maintain consistency across large numbers of cases.

Signal detection

Signal detection depends on finding meaningful patterns within safety data. If similar medical concepts are coded inconsistently, relevant cases may be distributed across different terms, making them harder to identify.

MedDRA’s structure also affects data retrieval. Its granularity and multiaxiality mean that related concepts may appear at different points in the hierarchy or under more than one SOC. These characteristics need to be considered when designing searches and interpreting results.

Aggregate reporting

Standardised coding supports the analysis and presentation of cumulative safety data. Consistent terminology makes it easier to identify patterns across cases and reporting periods.

Regulatory communication

Consistent medical terminology also facilitates communication between pharmaceutical companies, regulators, and other organisations that exchange safety information.

MedDRA Multiaxiality and Safety Data Analysis

MedDRA is not simply a list of terms organised by body system.

A PT can be linked to more than one SOC. One SOC is designated as the primary SOC, while other applicable assignments are secondary. This is known as multiaxiality.

This structure provides flexibility for medical analysis but also creates an important consideration for PV professionals.

Searching for only one SOC may not capture all relevant concepts.

For example, clinically related terms may be located in different SOCs, and some SOCs are not multiaxial. MedDRA guidance, therefore, recommends considering the structure of the terminology when designing data retrieval strategies.

This is one reason why MedDRA expertise matters beyond the initial coding step. The way data is coded directly affects how it can later be retrieved and interpreted.

Common MedDRA Coding Challenges

In real-world pharmacovigilance, coding decisions are not always straightforward.

Some of the most common challenges include:

Incomplete information.

The report may contain only a vague description of the event, making it difficult to select a highly specific term.

Ambiguous medical language.

Patients and healthcare professionals may use the same word differently or use non-standard terminology.

Symptoms versus diagnoses.

The report may describe symptoms without confirming an underlying disease.

Laboratory findings.

An abnormal test result should not automatically be converted into a clinical diagnosis.

Multiple concepts in one statement.

A single sentence may contain several distinct medical concepts that require separate consideration.

Different coding practices.

Different teams may interpret the same wording differently if coding conventions are not clearly established.

Changes between MedDRA versions.

The terminology is continuously maintained and updated, so organisations need appropriate version-management processes.

These challenges make training, documented conventions, and quality control important components of a reliable coding process.

MedDRA Version Management

MedDRA is continuously maintained and updated. The ICH M1 Points to Consider working group maintains guidance covering both term selection and data retrieval and presentation, with these documents updated annually alongside the March MedDRA release.

For organisations using MedDRA across long-running pharmacovigilance activities, version management is therefore an important operational consideration.

PV teams should know which MedDRA version is being used, understand how terminology changes may affect their processes, and apply appropriate procedures when transitioning between versions.

This becomes particularly relevant when organisations analyse historical safety data or compare information collected over different periods.

How to Improve MedDRA Coding Quality

High-quality MedDRA coding does not depend only on the experience of individual coders. It also requires a consistent process.

Organisations can improve coding quality by:

• establishing clear internal coding conventions;
• training coders on MedDRA structure and term selection principles;
• using qualified medical review when the clinical context is complex;
• performing quality checks on coded cases;
• documenting coding decisions and relevant conventions;
• maintaining control over MedDRA versions;
• regularly reviewing recurring coding inconsistencies.

The goal should be consistency without losing the meaning of the original safety information.

MedDRA also provides training and educational resources for users, including courses covering coding, data analysis, and other aspects of terminology use.

MedDRA Coding Is More Than Finding a Term

In pharmacovigilance, coding is often viewed as a routine step in case processing. In reality, it can have a much broader impact on the quality and usability of safety data.

A coding decision determines how reported medical information is represented in the safety database. That information may later contribute to case review, aggregate analysis, signal detection, regulatory reporting, and safety decision-making.

For this reason, effective MedDRA coding requires more than knowing how to search the terminology.

It requires understanding the reported medical concept, applying consistent principles, recognising the limitations of the available information, and considering how coded data will be used later.

When these principles are applied consistently, MedDRA helps transform diverse clinical descriptions into structured safety data that PV teams can analyse more reliably.