Watch the 4-minute video
4:20 · English · Fictional examples. The video loads from YouTube only when you press play. Watch on YouTube
Insight records often contain questions that nobody in the team can yet answer. They appear as a clinician’s query about a subgroup, a comment that local data are missing, or two regions interpreting the same issue differently. Unless someone looks for them deliberately, these questions stay scattered across meeting notes and insight entries and are never followed up.
The MAPS insights white paper defines an insight as new information, understanding, ideas or perspectives that may identify a gap, inform strategy or confirm existing views. [1] This article describes a practical way to find unanswered scientific questions in an insight database, check whether they are genuinely unanswered, describe why the evidence is missing and track each one until it is resolved with evidence. The companion Unresolved Scientific Question Log provides the register for that last step.
Why unanswered questions get lost
Questions hide inside other records
A question is rarely entered as a question. It may be recorded as an observation, a request, part of an interpretation or a follow-up note. The same uncertainty may be worded differently by different colleagues, in different countries and months apart. Searching only for the word “question” finds very little.
Recording an insight does not answer it
An insight record can show that a question exists. It does not show that the question matters to many clinicians, that no evidence exists, or that the team has addressed it. A field insight can identify a gap but does not itself resolve or validate it. Treat each candidate question as a hypothesis about missing evidence until it has been checked.
Find candidate questions in the insight database
Search for the signals of an open question
Review the insight database for the signals below. The example wording is fictional and shows the kind of phrasing to look for.
| Signal | What to look for | Fictional example wording |
|---|---|---|
| Direct question | A question the stakeholder asked that was not answered during the discussion. | “Is there evidence on outcomes in patients who switch treatment early?” |
| Stated uncertainty | Phrases such as “no data”, “unclear”, “we don’t know” or “not studied here”. | “Nobody seems to know how many patients reach a specialist within a month.” |
| Conflicting interpretation | Two records that read the same issue in different ways, with no evidence to decide between them. | One region attributes delays to referral rules; another attributes them to test availability. |
| Repeated theme without a decision | The same topic appears in several reviewed insights, but no review decision or action is recorded. | Pathway variation is mentioned in four insights over two quarters; none has a decision. |
| Unanswered follow-up | An action asked someone to “check the evidence”, but no result evidence was recorded. | Action closed as Completed with the result field left blank. |
A formal medical information request is not an insight. Route it through the approved medical information process, even if it also suggests an evidence gap. The gap, written without identifying details, can then be logged separately.
Group related records before counting them
Combine records that describe the same uncertainty, but keep a reference to each source record so the context can be recovered. Note whether the records come from one meeting, one professional network or several independent settings. The number of records in an insight database shows how often a topic was recorded. It is not an estimate of how common the issue is among clinicians or patients, because the database has no sampling frame.
Check that the question is really unanswered
Write the question precisely
A vague concern cannot be checked against the literature. Rewrite each candidate as a specific question. For questions about the effects of an intervention, the PICO elements are a useful structure: population, intervention, comparison and outcome. The Cochrane Handbook notes that well-formulated questions guide eligibility, searching, data collection and synthesis. [2] Other questions, such as those about care pathways or diagnostic workflow, still benefit from naming the population, setting and outcome of interest.
Look for existing reliable evidence first
The James Lind Alliance treats a treatment question as uncertain only when it cannot be resolved by reference to reliable, up-to-date systematic reviews of existing research. [3] A similar discipline helps in Medical Affairs. Before logging a question as a gap, check current systematic reviews, clinical guidelines, the approved product information and the organisation’s approved medical information responses. Record what was checked and when in Current evidence / source.
If the answer already exists, the need is different. It may be a communication or education need rather than an evidence gap, and it belongs in the relevant medical plan rather than the question log.
Describe why the evidence is missing
A framework developed for the Agency for Healthcare Research and Quality characterises research gaps using PICOS elements and classifies the reason a gap exists into four groups. [4,5] Naming the reason helps the team choose a proportionate next step. The possible approaches in the third column are suggestions for discussion, not part of the published framework.
| Reason the gap exists [4,5] | What it means in practice | Possible approach to discuss |
|---|---|---|
| Insufficient or imprecise information | Few or no studies, or results too imprecise to support a conclusion. | Structured literature review; consider whether new evidence generation is justified. |
| Biased information | Studies exist, but limitations in design or conduct make the results unreliable for this question. | Critical appraisal of the existing studies before planning anything new. |
| Inconsistent or unknown consistency | Results disagree, or there is only one study so consistency cannot be judged. | Compare study populations, settings and methods; seek expert scientific exchange on the differences. |
| Not the right information | Evidence exists, but not for the population, setting, comparison or outcome that matters here. | Check whether local or setting-specific evidence, such as a needs assessment, would answer the question. |
Record each question in the log
Enter each checked question as one row in the Unresolved Scientific Question Log. The fields match the Evidence Gaps tab of the Stakeholder Strategy and Engagement System, so records can be moved into the full system later. The example below is the first fictional demonstration row in the workbook.
| Field | What to record | Fictional example (GAP-001) |
|---|---|---|
| Gap ID | A unique, stable identifier. | GAP-001 |
| Strategy ID | Optional in the log: the medical objective the question supports. | STR-001 |
| Medical question | The precise question, as rewritten above. | Where does pathway variation occur? |
| Current evidence / source | What was checked, and its source and date. | Fictional literature review |
| Gap description | What is missing and, where useful, why. | Local pathway information is incomplete |
| Stakeholder need | Who needs the answer and for what purpose. | Specialists need pathway context |
| Priority | High, Medium or Low, as a documented team judgement. | High |
| Proposed evidence approach | The proportionate next step. | Structured needs assessment |
| Owner and review date | One accountable person and the next review. | Medical Lead A, 1 December 2026 |
| Status and resolution evidence | Open, In progress, Resolved or Deferred, with evidence when resolved. | Open; no resolution evidence yet |
Priority is a judgement, not a validated score. Consider the relevance to the medical objective, the consequence of leaving the question unanswered, how much uncertainty remains and the effort needed to address it. Write the reasoning in the gap description or proposed approach so another reviewer can follow it.
Review and resolve questions with evidence
Review the log on a regular cycle
Agree a review rhythm, such as monthly or quarterly, and an owner for the log itself. The Summary tab shows how many questions are unresolved and how many are past their review date. At each review, update the status and the proposed approach, and record any new evidence found.
Mark a question resolved only when evidence answers it
Completing an engagement, holding an advisory discussion or closing an action does not by itself answer the question. Mark a question Resolved only when the answering evidence is recorded in Resolution evidence and has been appropriately reviewed. The log flags any row marked Resolved without that evidence. Use Deferred, with a reason, when the team decides not to pursue a question now.
Return the result to the people who raised it
The MAPS white paper describes a feedback loop between those working with external stakeholders and those setting strategy. [1] When a question is resolved or deferred, tell the field team what was found or decided. This shows that reporting questions leads somewhere and improves the quality of future insight records.
Keep safety, requests and identifiers out of the log
The MAPS white paper places adverse event reporting and data privacy within the governance of insight work. [1] Do not record patient identifiers, adverse event or product complaint details, or confidential company information in the log. Handle those through the organisation’s approved processes and systems.
Before marking a question Resolved
• Is the question written precisely enough to be answered?
• Was existing reliable evidence checked first, and is the check recorded?
• Does the resolution evidence actually answer the question as written?
• Has the evidence been reviewed by an appropriate person?
• Has the result been returned to the team that raised the question?
References
1. McLoughlin M, Jonkman M, Zivkov M. Building Medical Insights Capabilities in Medical Affairs Organizations. Medical Affairs Professional Society, Insights Focus Area Working Group. Undated white paper. Sections on the insight definition, process, feedback and governance.
2. Thomas J, Kneale D, McKenzie JE, Brennan SE, Bhaumik S. Chapter 2: Determining the scope of the review and the questions it will address. Last updated August 2023. In Cochrane Handbook for Systematic Reviews of Interventions, version 6.5. Cochrane, 2024.
Read the Cochrane methods chapter
3. Chalmers I, Atkinson P, Fenton M, Firkins L, Crowe S, Cowan K. Tackling treatment uncertainties together: the evolution of the James Lind Initiative, 2003–2013. Journal of the Royal Society of Medicine. 2013;106(12):482–491. doi:10.1177/0141076813493063.
4. Robinson KA, Saldanha IJ, McKoy NA. Frameworks for Determining Research Gaps During Systematic Reviews. Methods Future Research Needs Report No. 2. AHRQ Publication No. 11-EHC043-EF. Rockville, MD: Agency for Healthcare Research and Quality; June 2011.
5. Robinson KA. A framework to identifying and characterise research gaps from systematic reviews. Poster, Cochrane Colloquium, Québec City, 2013. Source of the four reason categories quoted in the table.
Sources accessed 7 October 2026. The references support the cited professional and methodological principles. The search signals, review steps, possible approaches and examples are proposed practical approaches and have not been validated. Companion file: Unresolved Scientific Question Log version 1.0, catalogue ID MA-040. Its columns, dropdown choices and fictional example rows match the Evidence Gaps tab of the Stakeholder Strategy and Engagement System version 3.1.
Unresolved Scientific Question Log
Explore the workbook, its instructions and its fictional examples before adapting it to your team.
View the matching workbook