When patients are asked to complete a Clinical Outcome Assessment (COA), whether this is an eDiary, a symptom checklist or a quality of life questionnaire, it might appear to be a straightforward enough task. But beneath each checkbox and rating scale lies a complex process requiring cognitive effort. This mental effort, known as cognitive burden, can affect the quality and reliability of the data collected. If the burden is too high, patients may respond inaccurately skip questions or disengage entirely. How COAs are designed can reduce or aggravate this cognitive burden.

Understanding cognitive burden is crucial for anyone involved in the development, implementation or interpretation of COAs. This article looks at what cognitive burden entails, what contributes to it and how COAs can be designed to reduce it.

What is Cognitive Burden?

Cognitive burden refers to the mental workload required for a respondent to understand and complete an assessment. In the context of COAs, it typically involves four key components:

1.       Comprehension: Interpreting the meaning of the question. This includes understanding vocabulary, sentence structure and overall clarity.

2.       Memory Recall: Retrieving relevant information from memory, often across a specific time frame, such as ‘in the past week’, ‘in the past four weeks’.

3.       Judgement Formation: Making a subjective decision based on recalled experiences.

4.       Response Mapping: Translating this judgement into a predefined format, such as a Likert or a numerical scale.

Even a short question can involve all four steps, For example:

“How often did you feel fatigued in the past week? Always, Very Often, Often, Sometimes, Rarely”

The participant is required to understand the term “fatigued”, recalling one week of experiences, decoding how frequent the feeling was and selecting an appropriate response like “often” or “sometimes”.

Levels of Cognitive Burden

Not all questions place equal demands on respondents. Here are some examples (invented by myself for illustrative purposes):

Low Cognitive Burden:

“Do you currently have a headache? Yes/No”

This offers a simple, immediate timeframe. It presents the patient with a familiar concept and a straightforward yes/no binary response.

Medium Cognitive Burden:

“In the past week, how often did you feel anxious?”

This requires short-term recall. It uses subjective interpretation, e.g. “anxious” and moderate response complexity, such as a 5-point frequency scale.

High Cognitive Burden:

“In the past two weeks, how much has pain interfered with your ability to complete social, work-related, and self-care activities that are important to you?”

This involves multiple layers: time frame, pain severity, activity types and personal values. It requires abstraction and prioritisation and demand responses on a numeric or complex scale.

High-burden items could cause slower response times, increase the skipping of items or inaccurate responses.

Different Cognitive Strategies

Different individuals approach COAs with different cognitive strategies. Some people naturally chunk or deconstruct the concepts contained within complex questions into smaller parts. Other patients may rely more heavily on emotional impressions or intuitive judgements which may be less consistent but still valid.

Respondents may also prioritise ease: instead of deeply evaluating each question, they may resort to ‘satisficing’, i.e. choosing a near-enough answer rather than the most accurate one. High-burden assessments make this more likely.

Some patients try to visualise events, while others rely on a general sense of how they have been feeling. The ability to recall specific events or evaluate frequency is not evenly distributed across patients. This can be especially relevant in patients dealing with pain, fatigue or cognitive impairments.

Why We Need to Consider the Cognitive Burden

Other than the fact that items with a high cognitive burden makes surveys harder to fill out, it also threatens the reliability and validity of the results. If questions are misunderstood or skipped, the dataset can become fragmented. Consider what this can lead to:

-            Bias in data collection, where only more educated or cognitively resilient patients provide full responses

-            Lower response rates, especially in older adults or non-native speakers

-            Misleading outcomes, if patients misjudge their symptoms or choose incorrect response options.

This flawed data could feed directly into decisions about treatment, resource allocation or clinical trial outcomes.

How to Reduce the Cognitive Burden in COA Design

COA developers can take several steps to make assessments more cognitively accessible without sacrificing depth:

1.       Use Clear, Simple Language

Do not use jargon, double negatives or abstract phrasing. Jargon can be specific to the condition in question or it can be jargon common in clinical studies that are never explained in simpler terms.

2.       Break Down Complex Questions

Instead of combining multiple domains into one item, such as social, emotional and physical, they could be separated:

-            How much has pain interfered with your social activities?

-            How much has pain interfered with your work tasks?

And not:

-            How much has pain interfered with your social activities and/or work tasks?

Ideally, one item should carry one concept.

3.       Provide Anchors or Examples

Abstract terms like “usual activities” or “daily routine”  benefit from examples tailored to patient populations: “Daily routine, such as getting dressed, cooking, or walking the dog”.

4.       Simplify Time Frames

Patients recall recent experiences more accurately. Shorter, more intuitive time frames, such as “yesterday”, “this morning”, may yield better data than abstract periods like “the past 14 days”.

5.       Pilot with Target Users

Assessments should always be tested with people who match the target demographic. Part of the pilot testing should be not only testing if they can complete the assessment, but observing how they experience it cognitively and emotionally.

The Cognitive Burden as a Silent Variable

The cognitive burden in Clinical Outcome Assessments is a silent variable. It is often invisible to researchers but strongly felt by patients. If a question is too complex, abstract or cognitively demanding, the data may not reflect the reality it is meant to capture.

Good survey design should not merely seek to collect answers, but to create an environment where people can comfortably report the truth. This means minimising confusion, reducing unnecessary mental effort, as well as respecting the limitations of memory, attention and comprehension.

In Clinical Outcome Assessments of the future, understanding and addressing the cognitive burden is not only a technical issue, but a matter of patient dignity and trust.

Thank you for reading,

Mark Gibson

Stanhope, United Kingdom, June 2025