# RWE analysis-data fitness and observability assessment

> **Editorial worksheet, not an official FDA, EMA, ENCePP, ISPE, or ISPOR form.**
> Complete before cohort construction or outcome modelling. It records the
> evidence behind a data-source decision; it does not score, certify, or make a
> data source fit for purpose.

**Study / program:** [enter]  
**Decision owner:** [enter]  
**Version / date:** [enter]  
**Question and estimand:** [enter target population, treatment strategies,
outcome, summary measure, and time horizon]

## 1. Dataset and provenance

| Field | Record the evidence, not an assertion |
|---|---|
| Source and steward | [database, owner, refresh cadence, contact] |
| Population and setting | [who can enter, geography, care setting, inclusion mechanism] |
| Observation period | [available dates; known changes in capture or coding] |
| Analytic extract lineage | [source tables → transformations → analytic dataset; version IDs] |
| Linkage, if any | [files linked, method, match rate, unmatched-record handling, linkage validation] |
| Access and audit trail | [who can reproduce the extract; log/version locations] |

## 2. Relevance: can the source answer this question?

| Required element | Operational definition proposed | Where captured | Coverage / timing evidence | Gap or misclassification risk | Decision and mitigation |
|---|---|---|---|---|---|
| Target population / eligibility | [enter] | [enter] | [enter] | [enter] | [accept / restrict / augment / stop] |
| Treatment / exposure | [enter] | [order, dispense, administration, claim, etc.] | [enter] | [enter] | [enter] |
| Comparator | [enter] | [enter] | [enter] | [enter] | [enter] |
| Outcome | [enter] | [enter] | [enter] | [enter] | [enter] |
| Confounders | [enter] | [enter] | [enter] | [enter] | [enter] |
| Follow-up / censoring | [enter] | [enrollment, encounters, death, feed lag] | [enter] | [enter] | [enter] |

## 3. Reliability: can a reviewer trust the captured values?

| Dimension | Evidence examined | Finding | Material limitation | Action / owner / due date |
|---|---|---|---|---|
| Completeness | [missingness, enrollment continuity, feed coverage] | [enter] | [enter] | [enter] |
| Conformance | [code-system/version and allowed-value checks] | [enter] | [enter] | [enter] |
| Plausibility | [range, chronology, duplicate and impossible-sequence checks] | [enter] | [enter] | [enter] |
| Accrual lag | [event-to-availability distribution and cutoff rule] | [enter] | [enter] | [enter] |
| Change over time | [coding, benefit, EHR/vendor, or process changes] | [enter] | [enter] | [enter] |
| Record linkage | [match rate, error assessment, selection mechanism] | [enter] | [enter] | [enter] |

## 4. Decision record

**Fitness conclusion:** [fit as specified / fit with stated restrictions / not fit]  
**Evidence that would reverse the decision:** [enter]  
**Required protocol/SAP changes:** [enter]  
**Approved by / date:** [enter]

## Source basis

- U.S. Food and Drug Administration. *Real-World Data: Assessing Electronic
  Health Records and Medical Claims Data To Support Regulatory Decision-Making
  for Drug and Biological Products*. Guidance for Industry, July 2024.
  https://www.fda.gov/media/152503/download
- Wang SV, Pottegård A, Crown W, et al. HARmonized Protocol Template to Enhance
  Reproducibility of hypothesis evaluating real-world evidence studies on
  treatment effects. *Pharmacoepidemiology and Drug Safety*. 2023;32:44-55.
  https://doi.org/10.1002/pds.5507