Worked specimen
A draft outline that looks complete, and is not
This is the specimen the free check loads. It is a plausible draft for a single-arm cell therapy study with an external control: the headings are all present, the commitments are not. The check scores it 36/100 and finds 6 contradictions, 5 absences and 15 partly evidenced checks.
The outline, as pasted
RWE-01 Protocol outline, draft 0.9 Sponsor: Northline Therapeutics (invented for this specimen) Asset: NLT-101, an autologous cell therapy Indication: relapsed or refractory large B-cell lymphoma 1. Background and rationale NLT-101 is being studied in patients with relapsed or refractory large B-cell lymphoma who have failed two or more lines of therapy. The current standard of care in this population is uncertain and outcomes remain poor. A single-arm study with an external control is proposed because randomisation is not considered feasible in this population. 2. Objective and estimand Primary objective: to describe the overall response rate in patients treated with NLT-101 in routine practice. The estimand is not yet defined. 3. Study design Single-arm, retrospective cohort of patients treated between January and December. Patients will be followed from the first infusion. 4. Data sources Two hospital networks and a specialty lymphoma registry. A feasibility review was completed in August. Known limitations include missing laboratory values, coding changes during the study period and limited representation of community practice. 5. Study population Adults aged 18 years or older with relapsed or refractory large B-cell lymphoma who received at least one infusion of NLT-101. Inclusion criteria: confirmed diagnosis, at least two prior lines of therapy, and adequate organ function. Exclusion criteria: prior allogeneic transplant, active central nervous system involvement. Informed consent will be handled by the treating site. 6. Endpoints Primary endpoint: overall response rate at three months, defined as complete or partial response by the treating physician. Secondary endpoints: overall survival, progression-free survival and duration of response. 7. Statistical analysis Descriptive analysis. Overall response rate with a 95 percent exact confidence interval. Overall survival by Kaplan-Meier. Statistical analysis plan: to be finalised after database lock. 8. Sensitivity analyses Sensitivity analyses will be determined once the primary result is available. Subgroup analyses by age, prior lines of therapy and histology will be exploratory and chosen after review of the data. 9. External control Concurrent patients drawn from the same registry. External control patients will be matched on age and sex. Confounding will be assessed. 10. Data governance A data use agreement with the registry is in negotiation. The sponsor has the right to use the data for research. Source records are held by the registry. Data will be converted to CDISC SDTM for the submission. 11. Verification and timeline The confirmatory study is planned, with the first patient expected next year. No source data verification visit has been scheduled.
The first findings
The statistical analysis plan is pre-specified, dated and locked before analysis
ContradictedStatistical analysis plan: to be finalised after database lock.7. Statistical analysis, line 31
What resolves it. Name the SAP version, its date and who approved it, and state that it was locked before the analysis or unblinding. Record any change after lock as a dated amendment with its reason.
The estimand, primary endpoint and time zero are defined in advance
ContradictedThe estimand is not yet defined.2. Objective and estimand, line 11
What resolves it. Write the estimand in one sentence, name the primary endpoint and its ascertainment window, and state the index date and the end of follow-up. Keep the wording identical in the protocol, the SAP and the report.
The confirmatory study is underway, with milestones, not only on paper
ContradictedThe confirmatory study is planned, with the first patient expected next year. No source data verification visit has been scheduled.11. Verification and timeline, line 44
What resolves it. State the first patient in date, the enrolment target and the target completion date, and name the milestones that will be reported. If the study has not started, say what is blocking it and when enrolment opens.
Source data verification is planned, and the records can be retrieved
ContradictedThe confirmatory study is planned, with the first patient expected next year. No source data verification visit has been scheduled.11. Verification and timeline, line 44
What resolves it. Name the sites or data partners, the share of records to be verified, the verification method, who performs it, and the contractual right to produce the records. Say where the records are held and for how long.
Sensitivity analyses are pre-specified, each with the assumption it tests
ContradictedSensitivity analyses will be determined once the primary result is available.8. Sensitivity analyses, line 34
What resolves it. List each sensitivity analysis, the assumption it varies, the values it uses and the result that would make the primary finding fragile. Say which one is the primary sensitivity analysis.
A data use agreement is in place, with its scope written down
ContradictedA data use agreement with the registry is in negotiation.10. Data governance, line 41
What resolves it. Name the agreement, its date, its term, the permitted uses and the limits on publication and onward transfer. Say who signed for each party and where the executed copy is held.
The protocol is pre-specified and shared with FDA before the study starts
Not in the outlineWhat resolves it. State the date the protocol was shared, the meeting or submission it went through, and the review division. If it has not been shared, say when it will be and what is still open.
Outcome and exposure definitions are validated, not just coded
Not in the outlineWhat resolves it. Name the code set and its version, state the validation study or chart review behind it, give the positive predictive value or agreement, and say what happens to a record that does not meet the definition.
The full reading is on the free check.