PRO Burden: Deciding Which Questionnaires a Trial Actually Needs

A trial needs the patient-reported outcome (PRO) questionnaires that answer its stated objectives with suitable measures and a feasible assessment schedule. Six questionnaires are neither automatically excessive nor automatically justified. For a draft battery containing a quality-of-life questionnaire, disease-specific module, daily diary, symptomatic adverse event item set, physical function scale, and treatment satisfaction form, begin by mapping each candidate to the information the trial needs.

Usually, the first candidates to reconsider are an additional function scale that supplies no distinct evidence, a daily diary without a question requiring daily observations, and a satisfaction form without a defined exploratory objective. Preserve disease symptoms, tolerability, and needed function coverage. A broad quality-of-life measure may still serve an important stakeholder objective despite overlap with other measures. These are conditional decisions: the protocol team must establish what would be lost before removing anything.

The three tests below are this article’s practical synthesis of guidance and methods literature, rather than a validated algorithm or an FDA approval checklist. Apply them to the combined battery and document the rationale for every addition, including exploratory assessments.

1. Objective and measurement: What concept and prespecified objective does each questionnaire serve, what score will be used, and what supports its use in this population? Explain any intentional overlap.

2. Population and combined effort: Can participants understand and complete the whole battery under realistic trial conditions? Assess length, complexity, delivery, relevance, and the cumulative visit workload.

3. Timing and information: Does each assessment occur when it can answer the research question? Justify frequency and recall period separately, including a pretreatment reference when assessing change.

flowchart TD
    A["List candidate questionnaires and objectives"] --> B{"Distinct information or justified overlap?"}
    B -->|No| C["Reconsider the assessment and document information lost"]
    B -->|Yes| D{"Combined effort feasible for this population?"}
    D -->|Uncertain| E["Obtain patient input and test the proposed administration"]
    D -->|Supported| F{"Timing answers the research question?"}
    F -->|No| G["Revise timing without assuming recall equivalence"]
    F -->|Yes| H["Document the battery and monitoring plan"]
    C --> A
    E --> A
    G --> A

A protocol can inherit questionnaires from an earlier study, acquire a measure requested for payer evidence, and add exploratory instruments proposed by different stakeholders. Review those requests together. Individually plausible measures can ask participants about the same experience repeatedly, while an apparently long battery may contain information that patients consider important.

Aiyegbusi and colleagues’ 2022 Nature Communications article reviews burden as a combination of rationale, measure selection, delivery, patient involvement, and perception. It cautions against treating questionnaire length as a sufficient burden indicator or allowing concern about burden to exclude important outcomes without examining the evidence. It is a methodological discussion, rather than a trial establishing a universal safe battery size. 2022 burden considerations.

As a counter-example, the ISPOR practitioner summary reports that Atkinson and colleagues administered 14 questionnaires containing 176–180 items to high-risk bladder cancer patients undergoing radical cystectomy and urinary diversion, with low reported response burden. That population-specific observation, reported here through a secondary summary, does not establish tolerability for another trial or prove why participants found it acceptable. ISPOR summary and its cited counter-example.

Use item totals to describe workload, then ask whether the questions are relevant, repetitive, understandable, and deliverable. Removing a questionnaire that captures a needed patient experience can weaken the evidence even while reducing the tally. A shorter battery is useful only if it preserves the information required for the study’s objectives.

Create a decision record before negotiating the assessment schedule. Start with the concept and intended use of the score; then identify the instrument, version, population, language, and administration conditions. A measure’s name or its use in previous trials does not establish fitness for the present purpose. Final PFDD Guidance 3 describes the evidence needed for a fit-for-purpose interpretation within the intended context. Final PFDD Guidance 3, October 2025.

Overlap is a review signal, rather than an automatic exclusion rule. Two pain measures may serve different purposes or capture different periods. A utility measure may be needed for a planned economic evaluation even when disease-specific items also cover pain and activity. Conversely, an additional scale retained solely because it is familiar needs a clearer rationale.

For FDA-regulated anticancer therapy trials intended to demonstrate effects on survival, tumor response, or progression, the October 2024 Core PRO guidance recommends collecting and separately analyzing disease-related symptoms, symptomatic adverse events, overall side-effect impact, physical function, and role function. It recommends careful consideration of additional outcomes to focus on meaningful, measurable information while limiting burden. These are five concepts, not a requirement for five separate questionnaires. FDA Core PRO guidance, Sections I and III.

That guidance is nonbinding, does not endorse its example instruments, and does not extend its core set to all therapeutic areas or supportive-care trials. Outside its scope, use the logic of explicit concept selection while establishing the relevant outcomes for the actual population and trial objectives. Patient-reported symptomatic adverse events complement the trial’s standard safety data; they do not replace it.

For the evidence review of a named measure, see fit-for-purpose COA selection; for identifying the concepts patients consider important, see content validity and PFDD Guidance 2. Keep the present decision focused on the whole questionnaire package.

The 2009 FDA PRO guidance identifies questionnaire length, formatting, small fonts, new instructions for each item, consulting records, inadequate completion time, and literacy demands among contributors to burden. Tolerable burden depends on assessment frequency and timing, cognition, illness severity, and treatment toxicity. It also identifies missing data and refusal as possible indications of inappropriate burden or items. FDA PRO guidance, Section III.D.8.

Final PFDD Guidance 3 warns that length or complexity, a long combined battery, or very frequent administration may produce fatigue that affects how well responses reflect disease or treatment impact. Cognitive interviews and usability testing can inform this assessment. Its language describes a risk to investigate, rather than an inevitable loss of validity for every long measure. Final PFDD Guidance 3, Sections IV.F.3–4.

ISOQOL’s published minimum standards include acceptable patient and investigator burden alongside measurement evidence and score interpretability. This supports evaluating burden as part of instrument selection. Reeve and colleagues, 2013, abstract.

For this review, test the proposed package rather than estimating feasibility from one form in isolation. Invite people who reflect the intended trial population to work through the administration conditions, including its language, instructions, device or paper format, and position among other procedures. Record where participants pause, request assistance, struggle to interpret a response, or describe a question as irrelevant.

• Measure completion time in context, including interruptions and assistance. Use it to identify a problem and evaluate changes, without presenting a fixed number of minutes as a regulatory standard.

• Ask whether the purpose and questions are understandable and relevant. Document feedback across the intended range of functional and cognitive abilities rather than relying only on staff estimates.

• Inspect changes between instructions, response formats, and recall periods. Preserve the measure’s intended administration; a burden concern alone does not justify rewriting its questions.

• Review the entire visit and remote workload. A manageable questionnaire may be harder to complete after other demanding procedures or during periods of illness.

Patient input supports a feasibility judgment, but a favorable small pilot does not establish that the schedule will remain acceptable throughout treatment. State which participants and conditions were assessed, what changed, and what remains uncertain. Carry those uncertainties into a monitoring plan.

The Core PRO guidance recommends a pretreatment baseline, higher assessment frequency in the first few treatment cycles, consideration of the administration schedule, and different frequencies where the concepts and research objectives warrant them. Its example schedule is illustrative. Select the actual frequency using the trial’s objectives and expected disease and treatment trajectories. FDA Core PRO guidance, Section V.A.

Aiyegbusi and colleagues’ 2024 Nature Medicine consensus statement contains 19 recommendations developed through a literature review, two-stage Delphi survey, and consensus meeting. Its scheduling recommendations consider the disease trajectory, combined burden, and essential information. This is stakeholder consensus, not experimental proof of one optimal frequency or a regulatory requirement. 2024 respondent-burden consensus.

For each concept, work backward from the decision the data must support. If the question concerns a change at a landmark, establish the necessary baseline and follow-up observations. If the question concerns fluctuating symptoms around administration, determine whether the planned collection can observe that fluctuation in both treatment arms. A convenient cycle-start visit may miss information between visits.

Assessment frequency and recall period are separate choices. Administering a weekly-recall measure every three weeks does not make it a three-week measure. Removing a daily diary may discard within-week information even if another questionnaire asks about the same symptom. The 2009 guidance treats recall and instrument modifications as evidence questions; see PRO recall periods for that distinct decision. FDA PRO guidance.

Document the observation period each measure covers and any gaps the schedule deliberately leaves. A reduced frequency is defensible when it still addresses the objective, with an explanation of the information sacrificed. Preserve the planned endpoint observations rather than moving them solely to accommodate the rest of the battery.

Consider a hypothetical Phase 2 anticancer trial with a tumor-response objective and a key secondary disease-symptom endpoint. The draft package contains five visit questionnaires and one daily diary. All form lengths and decisions below are illustrative; they do not describe a real trial or prescribe modifications to named instruments. Assume the team will establish adequate evidence for the retained measures before using them.

The visit forms contain 30, 25, 12, 10, and 14 items: 30 + 25 + 12 + 10 + 14 = 91 items when all five are due. A separate five-item daily diary adds 5 × 7 = 35 responses over seven days. Keep these units separate: neither figure is a burden threshold, and the diary is not another five-item form administered only at a clinic visit.

For this example, assume broad health-related quality of life remains a documented descriptive secondary objective; the generic questionnaire also provides the needed physical and role function scores. The daily diary has no distinct short-term objective, the standalone function form adds no required information, and treatment satisfaction has no planned study question. Those assumptions, rather than the number six, determine the following decisions.

If the three retained sets and the added summary item are all due at one assessment point, the illustrative total becomes 30 + 25 + 12 + 1 = 68 items, a reduction of 23 from the original visit total. The daily diary is also removed. Actual effort still depends on administration frequency and conditions; a weekly tolerability schedule could add observations elsewhere. This arithmetic demonstrates a decision record, not a proven improvement in completion or a guarantee of regulatory conformance.

If pilot work shows that the retained package is still difficult, reconsider objectives, delivery, and supported modular options. Do not force the total lower by removing baseline, deleting required scoring items, or assuming that a global summary replaces symptom detail. The reduced package must still earn its place under all three tests.

Now suppose the hypothetical team keeps the original package without testing combined burden. Later in follow-up, participants leave forms unfinished and report difficulty completing them when unwell. This is an illustrative failure scenario, not an observed trial result or a prediction that every six-questionnaire battery will fail.

Draft PFDD Guidance 4, Section IV.A.7, describes risks of unaddressed burden or fatigue: more missing data, poorer responses such as repeatedly choosing the first option without considering the items, and more dropout. It also discusses patient input and pilot testing. As of this review, Guidance 4 remains a draft marked not for implementation; these considerations are not finalized requirements. Draft PFDD Guidance 4, April 2023.

Operationally, the important question is whether missing or rushed responses impair the intended endpoint observations. A repeated response pattern alone does not establish that answers are invalid. Similarly, missing assessments during illness do not by themselves establish a particular statistical missing-data mechanism. Record reasons and timing, and evaluate what assumptions the planned analysis can support.

A proposed monitoring approach is to have data management summarize expected and received assessments by instrument and timepoint, the outcome-assessment lead review administration concerns, clinical operations investigate participant and site obstacles, and biostatistics assess consequences for interpretation. Define who was expected to complete each assessment before calculating a completion rate. Review patterns across treatment arms and periods of high symptom burden without prompting participants toward particular responses.

The Core PRO guidance recommends prospective missing-data mitigation procedures, documented reasons in the analysis dataset, and a prespecified analysis plan. Use those records to distinguish scheduling or technical problems from possible burden-related loss. The credibility of a labeling claim depends on trial design, conduct, instrument suitability, and data quality; assess the consequences of missing observations in that context. FDA Core PRO guidance, Sections V.B and VI.

If the illustrative trial discovers that observations at its key symptom timepoint are disproportionately absent, a larger volume of exploratory data elsewhere does not replace them. The team must assess the credibility of the remaining evidence and its analysis assumptions. Preventing this loss is a reason to examine the whole package before the protocol is finalized.

A modular approach administers selected subscales rather than a complete multi-domain questionnaire. Serrano and colleagues’ 2024 position statement argues that this can be appropriate when selection is rigorous and supported by rationale and evidence. It is an expert position statement, not a controlled study proving that every shortened instrument preserves validity or reduces burden. Serrano and colleagues’ modular position statement.

The 2025 perspective by Ng and colleagues presents arguments for and against modular use. It highlights reduced comparability with studies administering full measures, potential item-order effects, uncertainty about acceptability to health technology assessment bodies, and the possibility of overlooking unexpected effects or favoring selected domains. These are tradeoffs to assess in context. Removing a subscale does not automatically invalidate every retained score, but a score requiring omitted items cannot simply be reconstructed as if the full instrument had been administered. Ng and colleagues’ 2025 perspective.

Distinguish selecting supported subscales from deleting individual questions inside a validated scoring structure. For any proposed change, identify the exact retained scores and administration evidence, and confirm the intended use of the instrument materials. The burden argument explains why a change is considered; it does not establish the change’s measurement properties.

For symptomatic adverse events, the Core PRO guidance recommends a concise selection based on mechanism, early data, and patient and provider input, including symptoms expected with either regimen in an active-control trial. It suggests considering a free-text question to avoid missing important symptoms. Justify the selected items for the trial. An overall side-effect impact summary adds a different perspective and does not substitute for symptom-specific measurement. FDA Core PRO guidance, Section IV.

Another option is to change when a justified measure is administered rather than changing its content. This can preserve required scores while avoiding unnecessary repetition, but fewer observations may leave important periods unmeasured. Explain that tradeoff in the protocol instead of treating fewer administrations as automatically equivalent evidence.

ICH E8(R1) recommends operationally feasible designs, avoidance of unnecessary complexity and data collection, and early involvement of patients and site staff who can identify overly burdensome procedures. It provides general design principles rather than a PRO-specific questionnaire limit. ICH E8(R1), Sections 3.3 and 3.3.3.

Turn the review into a usable record. For every retained assessment, state the objective, concept, instrument/version, score, evidence, timing, and responsible team. For each cut, state which information is no longer collected and why it is unnecessary for the planned objectives. For each unresolved modification, specify the evidence needed before implementing it.

• Protocol: carry the rationale into the schedule of assessments, with separate treatment of endpoint and exploratory purposes. Include procedures for avoiding and investigating missing assessments.

• Analysis plan: align the planned endpoint observations, score interpretation, and missing-data handling with the retained package. Explicitly acknowledge information that the revised collection will not provide.

• Feasibility record: document who reviewed the combined package, which administration conditions were tested, what problems emerged, and what changes followed.

• Instrument materials and operations: record the exact versions, selected scores, administration instructions, and any deviations from scoring manuals, with their rationale. Check the permissions applicable to the proposed implementation.

Use the related guides on protocol endpoint definitions, COA endpoint construction under draft PFDD Guidance 4, and COA licensing before an eCOA build for those downstream tasks. The battery review should leave the team able to explain every questionnaire it will administer and the information it intentionally chose to forgo.