Decision Questions and Evidence Strategy | Foundational | Source-pack reference date: November 5, 2026
Real-world data describe professional certification and health status. Real-world evidence is clinical evidence generated by analyzing those data. Your first task is to identify what an output actually establishes, how treatment was assigned, and whose decision it is intended to inform.
What you will learn
Distinguish real-world data, clinical evidence, and feasibility findings: Decision Questions and Evidence Strategy, Regulatory and payer evidence uses, Knowledge 1.
Classify data collection and treatment assignment independently: Regulatory and payer evidence uses, Knowledge 3.
Explain the section 505F program context for new indications and postapproval studies without inferring a lower approval standard: Regulatory and payer evidence uses, Knowledge 4.
Compare FDA questions with payer questions about comparative outcomes, utilization, patient experience, and population relevance: Regulatory and payer evidence uses, Knowledge 2.
Separate the course's U.S. FDA and Medicare focus from NICE's UK context and recognize the provider-defined certificate: Regulatory and payer evidence uses, Knowledge 5.
Estimated time
20-25 minutes for the lesson; allow additional time for the applied checks.
Blueprint mapping
Decision Questions and Evidence Strategy → Regulatory and payer evidence uses → parent domain weight 10%; no separate weight is supplied for this topic.
Prerequisites
None.
Key outputs
You should be able to choose an accurate data/evidence label, identify whether an output addresses feasibility or a clinical claim, classify assignment separately from data source, and route a research question to the appropriate decision context.
WHY
Why it matters
A large database can make a study possible without establishing that a treatment causes an outcome.
Routine professional certification records can support analyses within randomized or nonrandomized studies; the source does not determine assignment.
A regulatory evidence question and a coverage question may concern the same product while requiring different comparators, outcomes, and populations.
The supplied November 5, 2026 course blueprint places these distinctions before source selection, study design, and submission planning.
Where things stand: The supplied blueprint identifies section 505F as a program context involving new indications and postapproval studies. It also names FDA's August 2023 RWD/RWE guidance and December 2023 data-standards guidance in later operational requirements. This lesson uses that supplied context; it does not establish that a particular study is acceptable, exempt, or submission-ready.
Scope note: FDA, NICE, and ICH are named frameworks or organizations in the course context. Their inclusion does not establish that they issue or endorse this professional certificate. Completion criteria and assessment thresholds are provider-defined.
KEY TERMS
Key Terms & Must-Know Facts
Table 1. Foundational terms and decision boundaries
Term
Meaning in this lesson
Boundary to preserve
Real-world data, RWD
Data about health status or professional certification delivery routinely collected from sources such as electronic health records and claims.
Possessing records does not establish a clinical claim.
Real-world evidence, RWE
Clinical evidence about product use and potential benefits or risks generated through analysis of RWD.
The label does not guarantee adequacy for a particular decision.
Feasibility finding
An output about whether an intended study can be conducted, such as eligible counts or available follow-up.
It supports planning rather than automatically supporting effectiveness.
Clinical claim
A statement about a product's use, benefits, or risks.
The analysis must support the strength and scope of the statement.
Data source
The system or collection process from which study information originates.
It does not specify treatment assignment.
Treatment assignment
The process determining which treatment strategy a participant receives.
Randomized and nonrandomized assignment must be identified separately.
Randomization
Assignment through a chance-based study procedure.
Routine outcome capture does not remove randomization.
Nonrandomized study
A study without chance-based assignment by the study.
Analytical adjustment does not retrospectively randomize treatment.
Fit for purpose
Suitable for the specific question and intended decision.
Suitability is question-specific rather than a permanent database label.
Section 505F
The statutory program context highlighted here for RWE use involving new indications and postapproval studies.
It does not imply a lower approval standard.
Payer relevance
Relevance to coverage questions, covered populations, comparative outcomes, utilization, and patient experience.
Regulatory relevance does not establish every coverage question.
Transferability
The applicability of findings to another population or decision setting.
Shared product identity does not establish shared decision context.
Must-know facts:
RWD describe the inputs; RWE describes clinical evidence produced through analysis.
Counts of available people, records, and follow-up are feasibility outputs unless they answer an appropriately supported clinical question.
RWE can arise from randomized and nonrandomized studies.
Neither a familiar data source nor a sophisticated method establishes decision adequacy on its own.
Section 505F program relevance is distinct from acceptance of a particular claim.
FDA safety and effectiveness decisions differ from payer coverage questions, although some evidence needs overlap.
NICE belongs to the UK context; Medicare belongs to the U.S. payer context emphasized here.
CORE
Core content
1.1.1.1 1. Separate data, evidence, and feasibility
A Explanation: choose the conclusion the output supports
Decision conflict: A team wants to call a database extract proof of effectiveness. Choose a feasibility description when the output establishes available patients or fields; reserve a clinical conclusion for an analysis that addresses that clinical question with defensible methods.
WHY THIS IS TESTED: Mislabeling an early planning output can carry an unsupported claim into a protocol, evidence brief, or decision meeting.
Ask what was actually estimated. A count of potentially eligible patients addresses recruitment or analytical feasibility. A treatment-outcome analysis may generate RWE, but whether it supports a causal claim depends on design, measurement, bias, and uncertainty. Do not turn this into a rigid chronological rule: feasibility can also be reassessed during a study. The distinction is the question answered, not the file's creation date. See Fig 1.
Identify the proposed statement.
Inspect the actual output.
Separate availability from clinical inference.
Limit the conclusion to what was established.
Fig 1. Classify the output before the claim Exam takeaway: Patient counts establish availability, while clinical claims require question-matched analysis.
flowchart TD; A[Inspect output] --> B{Question answered}; B --> C[Availability or capture]; B --> D[Product use or clinical outcomes]; C --> E[Feasibility conclusion]; D --> F[Appraise analysis and claim support];
Routine records and analytical evidence Exam takeaway: Clinical evidence is generated through analysis rather than conferred by the presence of records.
{"prompt":"Square educational vector illustration on white with blue and gray palette and generous margins. Left: electronic health record screen and insurance claims file labeled Routine records. Center: analyst workstation labeled Analysis. Right: clinical evidence report labeled Findings and limitations. Use simple directional arrows, no numerical results, no implied approval stamp, no figure numbers.","caption":"Routine records and analytical evidence","size":"1024x1024"}
B Worked example: an impressive cohort count
A feasibility analyst searches records containing a 72-year-old woman with diabetes and chronic kidney disease. A mildly elevated cholesterol result appears in her chart. The network is nationally recognized, and its query returns 18,400 potentially eligible treatment users. The project lead proposes: 'The therapy improves outcomes in older adults.'
Identify task: Decide which statement the query supports.
Extract key facts: The output is a potentially eligible count; no outcome comparison has been reported. The age and conditions help describe the intended population.
Apply rule: Available records support feasibility, subject to verifying eligibility and required variables.
Eliminate distractors: Network reputation and the large count do not supply a missing treatment-outcome analysis; the cholesterol finding does not change the output's purpose.
Verify: Report potential availability and unresolved capture questions, then define the clinical question before developing a study.
C Exam trap
Wrong answer: The cohort size validates the effectiveness claim. Why tempting: A large denominator suggests precision and professional credibility. Single eliminating clue: The reported output contains no clinical outcome analysis.
D Checkpoint MCQ
A source-screening report lists potentially eligible patients, available treatment dates, and the proportion with follow-up. Which is the most appropriate conclusion?
A. The treatment has demonstrated comparative effectiveness.
B. The source has preliminary feasibility for the intended study.
C. The source is sufficiently reliable for any regulatory claim.
D. The target causal effect has been estimated with limited uncertainty.
Answer: B. A is tempting because treatment data are present, but no comparative outcome finding is provided. B is correct because the outputs address whether a study may be possible. C is tempting because availability resembles fitness, but reliability and claim-specific suitability remain to be assessed. D is tempting because follow-up supports estimation, but the report contains neither an effect estimate nor its uncertainty.
If you missed this: Review Separate data, evidence, and feasibility - focus on the question the output actually answers.
1.1.1.2 2. Classify source and treatment assignment independently
A Explanation: choose two labels instead of one
Decision conflict: A study uses routine electronic records, so a colleague labels it observational. Reject that shortcut. First identify how the relevant data were collected; then determine how treatment was assigned.
WHY THIS IS TESTED: Confusing these dimensions can obscure the study's design strengths and its remaining limitations.
A randomized study can capture outcomes through routine records. A nonrandomized study can use the same records to compare treatments selected during usual care. Conversely, dedicated study measurements do not establish random assignment. Studies may combine routine and dedicated data, so describe important components rather than forcing an inaccurate single-source label. Routine capture still requires assessment of outcome measurement and completeness. See Fig 2.
Table 2. Data collection and assignment are separate dimensions
Data collection
Randomized assignment
Nonrandomized assignment
Routine professional certification data
Study randomizes treatment; routine records capture relevant outcomes. This can generate RWE.
Usual-care treatment choices are analyzed using routine records. This can generate RWE.
Dedicated study collection
Study randomizes treatment and collects study-specific measurements; dedicated collection alone does not establish RWD status.
Study collects measurements without randomizing treatment; dedicated collection alone does not establish RWD status.
Mixed routine and dedicated collection
Describe randomization separately from each data component's role.
Describe nonrandomized assignment separately from each data component's role.
Identify each relevant data source.
Determine how treatment was assigned.
State both classifications.
Assess limitations of each dimension.
Fig 2. Two independent design questions Exam takeaway: Routine outcome capture can coexist with randomized treatment assignment.
flowchart TD; A[Study description] --> B[How were data collected]; A --> C[How was treatment assigned]; B --> D[Routine or dedicated or mixed]; C --> E[Randomized or nonrandomized]; D --> F[Report both dimensions]; E --> F;
Two-axis classification workspace Exam takeaway: A data-source label cannot substitute for a treatment-assignment label.
{"prompt":"Square blue and gray educational vector illustration with white margins. Show an analyst arranging a clean two by two board. Row labels: Routine data, Dedicated data. Column labels: Randomized assignment, Nonrandomized assignment. Place simple record and assignment icons in cells. Add a small Mixed sources card below the board. Do not label every cell RWE, do not show numerical results, and do not include figure numbers.","caption":"Two-axis classification workspace","size":"1024x1024"}
B Worked example: randomization with routine capture
A study enrolls adults with heart failure and diabetes, including a 69-year-old man whose chart also records stable mild anemia. The protocol randomly assigns two treatment strategies, while hospitalizations are identified through routine electronic records. Because no special hospitalization diary is collected, a reviewer calls the study nonrandomized.
Identify task: Classify assignment and outcome data collection.
Extract key facts: The protocol uses random assignment; hospitalization information comes from routine records.
Apply rule: This is randomized assignment with routine outcome capture, a possible route to RWE.
Eliminate distractors: Absence of a dedicated diary does not reverse randomization. Stable anemia is clinical noise for this classification task.
Verify: Confirm the randomization procedure and separately assess whether hospitalization capture is suitable for the question.
C Exam trap
Wrong answer: Prospective, study-specific measurements prove a study is randomized. Why tempting: Organized prospective collection resembles a conventional trial. Single eliminating clue: Treatment was selected by clinicians rather than assigned by chance.
D Checkpoint MCQ
A protocol randomly assigns treatment and obtains outcomes from routinely collected claims. Which description is most accurate?
A. Randomized assignment using routine data.
B. Nonrandomized assignment using routine data.
C. Randomized assignment using only dedicated study data.
D. Nonrandomized assignment using dedicated study data.
Answer: A. A is correct because it preserves both stated dimensions. B is tempting because claims often appear in observational research, but it contradicts random assignment. C is tempting because randomization suggests a conventional trial, but claims are described as routine. D is tempting if all study analyses are mistaken for dedicated collection, but it contradicts both facts.
If you missed this: Review Classify source and treatment assignment independently - focus on the assignment mechanism rather than the data format.
Quick Recall:
Fill in: Patient availability counts primarily address _____. Answer: feasibility.
True or false: Routine data require nonrandomized assignment. Answer: False.
What two labels should accompany a study description? Answer: Data collection/source and treatment assignment.
1.1.1.3 3. Separate program context from evidentiary adequacy
A Explanation: choose relevance without promising acceptance
Decision conflict: A sponsor identifies a new-indication question within the section 505F program context and assumes that routine records qualify automatically. Choose to assess whether the evidence can answer the question; reject automatic acceptance based on the program name.
WHY THIS IS TESTED: Program relevance identifies an evidence-planning opportunity, not a shortcut around evidence quality.
The supplied blueprint emphasizes new indications and postapproval studies. Explain those uses without claiming section 505F lowers the approval standard, mandates acceptance of an available source, or makes nonrandomized evidence sufficient in every circumstance. Match the proposed clinical statement to the study's actual support. Early advice and detailed FDA submission responsibilities are later topics; this foundation does not determine their outcome. See Fig 3.
Name the intended regulatory use.
Identify the relevant program context.
Assess question-specific evidence support.
State limitations without promising acceptance.
Fig 3. Program relevance versus evidence adequacy Exam takeaway: Section 505F context does not establish that a proposed clinical claim is adequately supported.
flowchart TD; A[New indication or postapproval question] --> B[Section 505F program context]; B --> C{Evidence supports intended claim}; C --> D[Explain support and limitations]; C --> E[Identify unresolved evidence gap]; D --> F[No automatic acceptance]; E --> F;
Evidence planning at a quality gate Exam takeaway: An RWE program pathway leaves the need for adequate supporting evidence in place.
{"prompt":"Square professional education vector image on white in blue and gray. Two document folders labeled New indication and Postapproval study approach a review desk labeled Evidence quality. A separate small sign reads Section 505F context. Show the reviewer examining methods and limitations, with no approval seal, no lower barrier, no claim that acceptance is guaranteed, no figure numbers, and generous margins.","caption":"Evidence planning at a quality gate","size":"1024x1024"}
B Worked example: a new-indication proposal
A team proposes a new-indication analysis using routine records that include a 66-year-old woman with chronic obstructive pulmonary disease and hypertension. Her latest blood pressure is mildly elevated. A competing product already has a prominent conference presentation. The sponsor argues that a section 505F-aligned project can proceed with a weaker comparator because the records already exist.
Identify task: Evaluate the sponsor's evidence-planning inference.
Extract key facts: The intended use is a new indication; comparator adequacy remains unresolved.
Apply rule: Program context does not lower the approval standard or establish the adequacy of this analysis.
Eliminate distractors: Existing records reduce neither the need for a suitable comparison nor the need to appraise bias. The blood pressure and competing presentation do not resolve that issue.
Verify: Document the intended claim and the comparator gap before representing the evidence as sufficient.
C Exam trap
Wrong answer: Postapproval status makes evidentiary limitations acceptable by default. Why tempting: Prior approval can be mistaken for assurance about every later question. Single eliminating clue: The current analysis addresses a distinct unresolved question.
D Checkpoint MCQ
Which statement best describes the section 505F context supplied in this course?
A. It makes existing routine data sufficient for new-indication claims.
B. It limits RWE to nonrandomized postapproval safety analyses.
C. It concerns RWE uses including new indications and postapproval studies without implying a lower approval standard.
D. It makes payer acceptance the criterion for regulatory evidence adequacy.
Answer: C. A is tempting because the program concerns RWE, but availability does not establish sufficiency. B is tempting because postapproval safety is a familiar use, but it improperly restricts both design and use. C correctly states the supplied scope and boundary. D is tempting because evidence can serve multiple audiences, but payer acceptance does not define regulatory adequacy.
If you missed this: Review Separate program context from evidentiary adequacy - focus on relevance versus sufficiency.
1.1.1.4 4. Match evidence to the decision-maker and jurisdiction
A Explanation: choose a decision question before an evidence brief
Decision conflict: A team wants one unchanged evidence conclusion for FDA, Medicare, and NICE. Choose a shared factual foundation with decision-specific interpretation; reject the assumption that the audiences ask identical questions.
WHY THIS IS TESTED: An internally sound study can leave the intended decision unanswered when its comparator, outcomes, or population are mismatched.
FDA safety and effectiveness questions concern the regulatory use under consideration. Payers may ask about comparative outcomes, utilization, patient experience, and relevance to covered populations. These interests can overlap: the distinction is emphasis and decision context, not an exclusive division in which regulators ignore patient experience or payers ignore safety. NICE provides a UK comparison; do not import its conclusions as U.S. Medicare coverage determinations. See Fig 4.
Table 3. Decision audiences and appropriate interpretation
Audience or context
Question emphasis
Interpretive boundary
U.S. FDA
Safety and effectiveness for the proposed regulatory use.
A regulatory finding does not answer every coverage question.
U.S. payer, including Medicare context
Comparative outcomes, utilization, patient experience, and relevance to covered populations.
Relevance must be established for the specific decision rather than presumed from product approval.
NICE, UK context
Appraisal within its UK decision setting.
A NICE conclusion is not a U.S. Medicare determination.
Certificate provider
Completion and assessment against course-defined criteria.
Named agencies are not thereby certificate issuers or endorsers.
Identify decision-maker and jurisdiction.
State the decision to be informed.
Check population, comparator, and outcomes.
Adapt the interpretation and disclose gaps.
Fig 4. One evidence base, distinct decision questions Exam takeaway: Evidence interpretation should follow the intended decision rather than the product name alone.
flowchart TD; A[Evidence base] --> B[U.S. FDA question]; A --> C[U.S. payer question]; A --> D[UK NICE context]; B --> E[Safety and effectiveness use]; C --> F[Comparative outcomes and covered population]; D --> G[UK appraisal relevance];
Paired regulatory and payer workstations Exam takeaway: Shared findings may need different comparator, outcome, and population explanations for different decisions.
{"prompt":"Square blue and gray flat educational illustration on white. Two equal workstations share the same central study report. Left workstation labeled FDA reviews cards Safety and Effectiveness. Right labeled Payer reviews cards Comparative outcomes, Utilization, Patient experience, Covered population. Show overlapping interests through the shared report rather than mutually exclusive evidence bins. No seals, invented numbers, approval claims, or figure numbers.","caption":"Paired regulatory and payer workstations","size":"1024x1024"}
B Worked example: a coverage brief with the wrong audience
A medical affairs analyst prepares a U.S. Medicare-facing brief for a population that includes a 76-year-old man with diabetes and osteoarthritis. His stable mildly reduced hemoglobin appears in the source summary. A polished UK NICE appraisal is available, and the product has an FDA regulatory history. The analyst proposes copying the UK conclusion as the U.S. coverage conclusion.
Identify task: Decide how to frame evidence for the intended payer audience.
Extract key facts: The target decision is in a U.S. Medicare context; the existing appraisal belongs to a UK context.
Apply rule: Assess relevant population, comparator, outcomes, utilization, and patient experience for the intended question.
Eliminate distractors: A polished appraisal and FDA history are useful context, but neither establishes the U.S. coverage conclusion. The hemoglobin finding does not determine jurisdictional relevance.
Verify: Identify what the shared evidence answers and what remains unresolved before drafting a decision-specific brief.
C Exam trap
Wrong answer: Regulatory approval resolves which comparator a payer should consider most relevant. Why tempting: Approval represents a consequential evidence decision. Single eliminating clue: The payer asks about a different covered-population treatment alternative.
D Checkpoint MCQ
A study supports a regulatory safety question, but a payer asks about utilization compared with a commonly used alternative. What is the most appropriate next step?
A. Treat the regulatory finding as sufficient for the utilization question.
B. Replace the regulatory conclusion with an unsupported comparative claim.
C. Use a UK appraisal conclusion as the U.S. coverage determination.
D. Specify the payer question and assess whether the available evidence addresses it.
Answer: D. A is tempting because the evidence already informed an important decision, but the outcomes differ. B is tempting because it produces an audience-specific statement, but the statement lacks supporting analysis. C is tempting because an appraisal is decision-focused, but the jurisdiction and decision are different. D is correct because it tests the evidence against the actual information need.
If you missed this: Review Match evidence to the decision-maker and jurisdiction - focus on the decision-specific evidence gap.
Quick Recall:
Does section 505F context establish acceptance of an analysis? Answer: No.
Which context does NICE represent here? Answer: UK appraisal.
Who defines this certificate's assessment thresholds? Answer: The course provider.
TRAPS
Exam Traps & Differentiators
Common traps & distractors:
Choose the largest database first. Tempting because scale appears to increase credibility; the eliminating clue is that the required outcome is not captured.
Call adjusted treatment groups randomized. Tempting because measured characteristics may look similar; the eliminating clue is that treatment was selected during usual care.
Use the same conclusion for every audience. Tempting because the product and study are unchanged; the eliminating clue is a different decision question.
Table 4. Looks similar but is not equivalent
Superficially similar statements
Discriminating question
Appropriate distinction
Many records versus persuasive clinical evidence
What clinical analysis supports the statement?
Volume and inferential support are separate.
Outcome association versus treatment effect
What supports a causal interpretation?
An observed difference alone does not establish causation.
Balanced measured characteristics versus randomization
How was treatment assigned?
Adjustment changes an analysis, not historical assignment.
Regulatory usefulness versus payer usefulness
Which decision and population are being addressed?
Usefulness must be assessed for the intended question.
Named agency versus credential issuer
Who defines completion requirements?
Course context does not establish agency endorsement.
Fig 5. Choose the narrower defensible conclusion Exam takeaway: When a proposed statement exceeds the analysis, narrow the statement or identify additional evidence needed.
flowchart TD; A[Proposed conclusion] --> B{Matches actual analysis}; B -->|Yes| C[State findings and limitations]; B -->|No| D[Reject overstatement]; D --> E[Narrow conclusion]; D --> F[Define additional evidence need];
Use the conclusion check in Fig 5 before finalizing an evidence brief.
Jurisdiction and decision routing Exam takeaway: A UK appraisal can inform discussion without becoming a U.S. coverage determination.
{"prompt":"Square educational vector illustration with white background and blue gray palette. Show two clearly separated document trays: United States with FDA and Medicare cards, and United Kingdom with NICE card. A neutral evidence report sits between trays with dotted contextual arrows, not equivalence arrows. Add small labels Decision and Population on each tray. Avoid official logos, flags, endorsements, numerical findings, and figure numbers.","caption":"Jurisdiction and decision routing","size":"1024x1024"}
If the stem says...
'The outcome is absent from the source' → question fitness before choosing an analytical method.
'Clinicians chose treatment' → classify assignment as nonrandomized even if later adjustment is extensive.
'The payer's usual comparator differs' → identify an unresolved coverage-relevance question.
'The course names FDA and NICE' → separate educational scope from credential endorsement.
What is coming next: The next topic, Decision Questions and Target Effects, turns these distinctions into an explicit population, intervention or exposure, comparator, outcomes, setting, follow-up, and target effect. The supplied material identifies no specific forthcoming rule or implementation date for this topic; do not infer a new requirement from this lesson.
ALGO
Algorithm / Approach
Use this five-step intake check before selecting a data source or drafting a claim. See Fig 6.
1. Name the decision. Identify the decision-maker, jurisdiction, and intended use.
2. Inspect the output. Distinguish source records, feasibility results, and clinical evidence.
3. Classify the study. State data collection and treatment assignment separately.
4. Match support to claim. Identify what the design and measurements can establish.
5. State the next evidence need. Document the supported conclusion and unresolved gap.
Fig 6. Operational evidence-intake algorithm Exam takeaway: Clarify the decision and the output before selecting methods or strengthening a claim.
flowchart TD; A[Name decision and jurisdiction] --> B[Classify available output]; B --> C[Separate source and assignment]; C --> D[Match evidence to proposed claim]; D --> E[Record conclusion and remaining gap];
Evidence-intake meeting Exam takeaway: A useful intake summary distinguishes what is known from what must be established next.
{"prompt":"Square professional education illustration, clean blue and gray vector style on white. Analyst and evidence reviewer sit at a simple desk with five clearly separated cards labeled Decision, Output, Source, Assignment, Gap. Behind the cards are a routine record icon and a report icon. Show a focused operational review with no patient identifiers, no treatment advice, no approval badge, no numerical results, and generous margins. No figure numbers.","caption":"Evidence-intake meeting","size":"1024x1024"}
Boundary check: Completing this intake does not establish research permission, applicable human-subject protections, IND status, safety-reporting arrangements, or submission readiness. The course addresses those determinations in later topics. Keep the methodological classification separate from those practical obligations.
QUIZ
Self-check quiz
Choose one best answer for each question. All scenarios are illustrative operational exercises.
1. Feasibility or effectiveness?
A source assessment identifies 9,600 potentially eligible patients and confirms that hospitalization fields are present. No comparison has been conducted. What is the strongest supported statement?
A. The treatment reduces hospitalization.
B. The source has preliminary availability relevant to the planned study.
C. The source has established causal validity for the intended claim.
D. The source supports the same clinical conclusion in every covered population.
2. Classify the design
Clinicians choose treatments during usual care. Investigators later analyze routine records and adjust measured baseline differences. Which description is correct?
A. Randomized assignment because the analysis balances baseline variables.
B. Randomized assignment because the data reflect usual practice.
C. Nonrandomized assignment using routine data.
D. Nonrandomized assignment using only dedicated study collection.
3. Interpret program relevance
A sponsor proposes RWE for a new indication and cites section 505F. The proposed outcome is poorly captured. What is the most appropriate response?
A. Program relevance permits the measurement limitation to be ignored.
B. Existing approval for another indication establishes this new claim.
C. Routine collection makes outcome validation unnecessary.
D. Address outcome suitability because program relevance does not establish evidence adequacy.
4. Match the audience
An evidence package addresses a regulatory safety question in a broad adult population. A payer asks about comparative utilization in its older covered population. What should the team do first?
A. Specify the payer question and assess the package against it.
B. Adopt the regulatory conclusion as the utilization conclusion.
C. Substitute a NICE conclusion for the U.S. payer decision.
D. Increase the source cohort size before defining the comparator.
5. Integrate source, design, and decision
A randomized study captures hospitalizations through routine records. A U.S. payer asks whether its population and comparator are represented. Which statement best integrates these facts?
A. Routine capture makes the study nonrandomized, so relevance is established through adjustment.
B. Randomization makes population and comparator relevance unnecessary to examine.
C. The study uses randomized assignment with routine data, while payer relevance needs separate assessment.
D. Routine outcome capture makes the findings applicable to any payer using the same diagnosis codes.
ANSWERS
Answer key & explanations
1. Answer: B
A: Tempting because hospitalization data are present, but field availability does not establish an outcome difference. B: Correct because eligibility counts and field presence support preliminary feasibility. C: Tempting because feasibility is necessary for many studies, but no causal analysis or validity assessment is supplied. D: Tempting because many patients may appear representative, but neither a clinical conclusion nor transferability has been established.
If you missed this: Review Separate data, evidence, and feasibility - focus on availability versus clinical inference.
2. Answer: C
A: Tempting because adjustment can make measured characteristics more comparable, but it does not change how treatment was assigned. B: Tempting because usual practice suggests broad applicability, but practice setting does not create randomization. C: Correct because clinicians selected treatments and routine records supplied the data. D: Tempting because the analysis is study-specific, but study-specific analysis is not dedicated data collection.
If you missed this: Review Classify source and treatment assignment independently - focus on assignment at the time treatment was selected.
3. Answer: D
A: Tempting because the intended use fits the named program context, but relevance does not excuse an unsuitable outcome measure. B: Tempting because prior approval indicates evidence exists for another use, but it does not establish the proposed new indication. C: Tempting because routine collection appears authentic, but routine records can still fail to measure the needed outcome adequately. D: Correct because adequacy must be evaluated for the proposed clinical question.
If you missed this: Review Separate program context from evidentiary adequacy - focus on the unchanged need for question-specific support.
4. Answer: A
A: Correct because the population, comparator, and utilization outcome define a potentially different evidence need. B: Tempting because the package already addresses a consequential regulatory question, but safety and utilization conclusions are not interchangeable. C: Tempting because NICE appraisals are decision-oriented, but UK conclusions are not U.S. payer determinations. D: Tempting because more records may improve some analyses, but size cannot resolve an undefined or mismatched question.
If you missed this: Review Match evidence to the decision-maker and jurisdiction - focus on defining the actual coverage question.
5. Answer: C
A: Tempting because routine records are commonly used in observational studies, but the scenario explicitly states randomization and does not establish relevance. B: Tempting because randomization strengthens the assignment process, but it does not guarantee relevance to a different population or comparator. C: Correct because it preserves the source/design distinction and identifies the payer's separate applicability question. D: Tempting because shared coding suggests comparable measurement, but coding similarity alone does not establish population or decision applicability.
If you missed this: Review Classify source and treatment assignment independently and Match evidence to the decision-maker and jurisdiction - focus on distinguishing design classification from decision relevance.
RAPID
Rapid review / cheat-sheet
Claim wording → do not let the verb imply more than the analysis establishes.
Source suitability → specify the intended question before declaring a source useful.
Mixed collection → describe the role of each important data component.
Routine capture → still examine whether the required outcome is measured adequately.
Measured balance → read it as an analytical property, not an assignment history.
Cross-audience reuse → preserve factual findings while rebuilding the decision-specific interpretation.
Evidence gap → state what remains unanswered rather than treating it as a negative treatment finding.
Study classification → keep it separate from research permissions and submission obligations.
Key points: Distinguish available data from supported conclusions; classify assignment separately from source; assess program relevance without promising acceptance; and match evidence to the intended audience and jurisdiction.
End of your free lesson
Keep going with the full Real-World Evidence for Regulatory and Payer Decisions Professional Certificate
You just studied the way members do, in the same lesson interface. Full access unlocks every lesson, the question bank and flashcards for a full year.