Why English? | Reporting Nonrandomised Public Health Evaluations With TREND Allocation, Intervention Exposure and Participant Flow

People searching for the TREND statement, nonrandomized evaluation reporting or public health intervention evaluation need more than a success story. English matters because the report must show how units reached each condition, what was delivered, who remained, what was measured and which differences could explain an observed association.

The CDC’s current TREND page, reviewed by subject-matter experts on 27 May 2025 and checked here on 5 November 2026, describes Transparent Reporting of Evaluations with Nonrandomized Designs as a 22-item checklist for standardised reporting of nonrandomised controlled trials. The official CDC checklist record preserves the 2004 statement for behavioural and public-health intervention evaluations.

TREND improves reporting; it does not make an evaluation randomised or remove confounding. This article owns nonrandomised public-health evaluation reporting. CONSORT owns randomised-trial reporting, TIDieR owns intervention description, and ROBINS-I owns risk of bias in nonrandomised intervention results. Keeping those purposes separate lets readers value a transparent report without promising that its programme will work identically in every school, clinic or community.

• Identify the unit and method of assignment.
• Describe intervention content, provider, setting and exposure.
• Follow participants from screening to analysis.
• Report baseline differences, outcomes, uncertainty and adverse effects.
• Interpret associations within the limits of a nonrandomised design.

---

Identify the Evaluation as Nonrandomised

The title and abstract should state how units reached intervention conditions so readers do not assume random allocation.

Reporting move for Identify the Evaluation as Nonrandomised

Reconstruct the evaluation as a timeline from recruitment through assignment, delivery, follow-up and analysis. A timeline exposes whether eligibility, intervention start and outcome measurement were aligned across conditions. Keep the heading **Identify the Evaluation as Nonrandomised** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

Two schools may both appear in the intervention group because they volunteered early. If those schools also have more experienced staff, the allocation story belongs beside the outcome, not buried in a limitation paragraph. For **identify the evaluation as nonrandomised**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **identify the evaluation as nonrandomised** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Use a Structured Abstract

Summarise population, design, allocation, intervention, outcomes and main limitations without hiding the nonrandomised design.

Reporting move for Use a Structured Abstract

Draw the units in nested boxes—community, organisation, class and person—and write the eligibility and assignment rule at each level. This prevents an individual-level conclusion from hiding a school-level allocation process. Keep the heading **Use a Structured Abstract** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A community campaign offered six sessions, but attendance ranged from zero to six. Reporting only assignment erases the difference between access and exposure; reporting only completers can create a selected comparison. For **use a structured abstract**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **use a structured abstract** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Explain the Scientific and Behavioural Rationale

Connect the public-health problem, relevant theory and proposed mechanism to the intervention components.

Reporting move for Explain the Scientific and Behavioural Rationale

Build an intervention exposure table with planned dose, offered dose, received dose and completed dose. Nonrandomised comparisons often become clearer when exposure is separated from mere assignment. Keep the heading **Explain the Scientific and Behavioural Rationale** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

If a policy starts in one district before another, calendar time may bring other changes. The report should show dates, concurrent programmes and measurement schedules before attributing every difference to the intervention. For **explain the scientific and behavioural rationale**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **explain the scientific and behavioural rationale** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Define Eligibility at Every Sampling Level

A community, school, clinic, class and participant can each have different entry criteria that affect interpretation.

Reporting move for Define Eligibility at Every Sampling Level

Copy the baseline characteristics that matter causally, then explain why each might affect both participation or programme choice and the outcome. Do not turn the table into a ritual significance test. Keep the heading **Define Eligibility at Every Sampling Level** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A flow diagram can reveal that participants with missing follow-up were more common in one condition. The next question is why they left and how the analysis handled information that may depend on outcome risk. For **define eligibility at every sampling level**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **define eligibility at every sampling level** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Describe Recruitment and Setting

Referral, self-selection, outreach and gatekeeper routes determine who had a realistic chance to participate.

Reporting move for Describe Recruitment and Setting

Follow every denominator. The number invited, enrolled, assigned, exposed, followed and analysed can differ; a percentage is meaningful only when its denominator and condition are visible. Keep the heading **Describe Recruitment and Setting** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A statistically adjusted result still depends on measured variables, their quality and the chosen model. The report should name the adjustment set and explain why it addresses the main design differences. For **describe recruitment and setting**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **describe recruitment and setting** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Describe Intervention Content and Delivery

Report what was delivered, by whom, to what grouping, in which setting, through which channel and over what period.

Reporting move for Describe Intervention Content and Delivery

Pair every outcome estimate with its time point, precision and analysis population. Then label adjusted, subgroup and sensitivity analyses as planned or exploratory so flexibility remains visible. Keep the heading **Describe Intervention Content and Delivery** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

An unintended effect may be burden, stigma, displacement or unequal access rather than a clinical adverse event. TREND’s public-health purpose makes those consequences worth reporting explicitly. For **describe intervention content and delivery**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **describe intervention content and delivery** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

State Objectives and Hypotheses

Specific objectives keep confirmatory claims separate from exploratory observations.

Reporting move for State Objectives and Hypotheses

End the section with a transfer sentence naming setting, staffing, participant mix and implementation conditions. This protects a useful local evaluation from becoming an unsupported universal promise. Keep the heading **State Objectives and Hypotheses** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

For a family reading a school programme evaluation, a responsible summary says which learners participated, how the comparison arose and what changed—not that every future learner will receive the same benefit. For **state objectives and hypotheses**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **state objectives and hypotheses** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Define Primary and Secondary Outcomes

Name instruments, collection methods, time points and data-quality procedures rather than promising to measure impact broadly.

Reporting move for Define Primary and Secondary Outcomes

Reconstruct the evaluation as a timeline from recruitment through assignment, delivery, follow-up and analysis. A timeline exposes whether eligibility, intervention start and outcome measurement were aligned across conditions. Keep the heading **Define Primary and Secondary Outcomes** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

Two schools may both appear in the intervention group because they volunteered early. If those schools also have more experienced staff, the allocation story belongs beside the outcome, not buried in a limitation paragraph. For **define primary and secondary outcomes**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **define primary and secondary outcomes** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Explain Sample-Size Decisions

State assumptions, target numbers and any interim or stopping rules that apply to the evaluation.

Reporting move for Explain Sample-Size Decisions

Draw the units in nested boxes—community, organisation, class and person—and write the eligibility and assignment rule at each level. This prevents an individual-level conclusion from hiding a school-level allocation process. Keep the heading **Explain Sample-Size Decisions** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A community campaign offered six sessions, but attendance ranged from zero to six. Reporting only assignment erases the difference between access and exposure; reporting only completers can create a selected comparison. For **explain sample-size decisions**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **explain sample-size decisions** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Name the Unit and Method of Assignment

Individuals, classrooms, clinics or communities may be assigned by policy, timing, geography, choice or another rule.

Reporting move for Name the Unit and Method of Assignment

Build an intervention exposure table with planned dose, offered dose, received dose and completed dose. Nonrandomised comparisons often become clearer when exposure is separated from mere assignment. Keep the heading **Name the Unit and Method of Assignment** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

If a policy starts in one district before another, calendar time may bring other changes. The report should show dates, concurrent programmes and measurement schedules before attributing every difference to the intervention. For **name the unit and method of assignment**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **name the unit and method of assignment** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Report Methods Used to Reduce Allocation Bias

Matching, stratification or adjustment should be described with variables, timing and limitations.

Reporting move for Report Methods Used to Reduce Allocation Bias

Copy the baseline characteristics that matter causally, then explain why each might affect both participation or programme choice and the outcome. Do not turn the table into a ritual significance test. Keep the heading **Report Methods Used to Reduce Allocation Bias** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A flow diagram can reveal that participants with missing follow-up were more common in one condition. The next question is why they left and how the analysis handled information that may depend on outcome risk. For **report methods used to reduce allocation bias**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **report methods used to reduce allocation bias** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Explain Blinding Where It Was Possible

State who knew the condition and how awareness could influence behaviour, measurement or analysis.

Reporting move for Explain Blinding Where It Was Possible

Follow every denominator. The number invited, enrolled, assigned, exposed, followed and analysed can differ; a percentage is meaningful only when its denominator and condition are visible. Keep the heading **Explain Blinding Where It Was Possible** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A statistically adjusted result still depends on measured variables, their quality and the chosen model. The report should name the adjustment set and explain why it addresses the main design differences. For **explain blinding where it was possible**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **explain blinding where it was possible** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Use Statistical Methods That Match the Design

Account for clustering, repeated measures, baseline differences, missing data and prespecified adjusted analyses.

Reporting move for Use Statistical Methods That Match the Design

Pair every outcome estimate with its time point, precision and analysis population. Then label adjusted, subgroup and sensitivity analyses as planned or exploratory so flexibility remains visible. Keep the heading **Use Statistical Methods That Match the Design** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

An unintended effect may be burden, stigma, displacement or unequal access rather than a clinical adverse event. TREND’s public-health purpose makes those consequences worth reporting explicitly. For **use statistical methods that match the design**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **use statistical methods that match the design** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Draw Participant Flow Through Every Stage

Show screening, eligibility, enrolment, assignment, intervention exposure, follow-up and analysis for each condition.

Reporting move for Draw Participant Flow Through Every Stage

End the section with a transfer sentence naming setting, staffing, participant mix and implementation conditions. This protects a useful local evaluation from becoming an unsupported universal promise. Keep the heading **Draw Participant Flow Through Every Stage** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

For a family reading a school programme evaluation, a responsible summary says which learners participated, how the comparison arose and what changed—not that every future learner will receive the same benefit. For **draw participant flow through every stage**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **draw participant flow through every stage** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Report Protocol Deviations

Name departures from planned recruitment, assignment, delivery or analysis and explain when and why they occurred.

Reporting move for Report Protocol Deviations

Reconstruct the evaluation as a timeline from recruitment through assignment, delivery, follow-up and analysis. A timeline exposes whether eligibility, intervention start and outcome measurement were aligned across conditions. Keep the heading **Report Protocol Deviations** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

Two schools may both appear in the intervention group because they volunteered early. If those schools also have more experienced staff, the allocation story belongs beside the outcome, not buried in a limitation paragraph. For **report protocol deviations**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **report protocol deviations** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Give Baseline Characteristics by Condition

Readers need to see whether groups differed before the intervention and which differences matter substantively.

Reporting move for Give Baseline Characteristics by Condition

Draw the units in nested boxes—community, organisation, class and person—and write the eligibility and assignment rule at each level. This prevents an individual-level conclusion from hiding a school-level allocation process. Keep the heading **Give Baseline Characteristics by Condition** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A community campaign offered six sessions, but attendance ranged from zero to six. Reporting only assignment erases the difference between access and exposure; reporting only completers can create a selected comparison. For **give baseline characteristics by condition**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **give baseline characteristics by condition** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

State the Denominator for Every Result

Keep enrolled, exposed, followed and analysed populations distinct so percentages remain interpretable.

Reporting move for State the Denominator for Every Result

Build an intervention exposure table with planned dose, offered dose, received dose and completed dose. Nonrandomised comparisons often become clearer when exposure is separated from mere assignment. Keep the heading **State the Denominator for Every Result** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

If a policy starts in one district before another, calendar time may bring other changes. The report should show dates, concurrent programmes and measurement schedules before attributing every difference to the intervention. For **state the denominator for every result**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **state the denominator for every result** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Report Outcomes With Precision

Present estimates and uncertainty for each condition and comparison, not only P values or verbal success labels.

Reporting move for Report Outcomes With Precision

Copy the baseline characteristics that matter causally, then explain why each might affect both participation or programme choice and the outcome. Do not turn the table into a ritual significance test. Keep the heading **Report Outcomes With Precision** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A flow diagram can reveal that participants with missing follow-up were more common in one condition. The next question is why they left and how the analysis handled information that may depend on outcome risk. For **report outcomes with precision**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **report outcomes with precision** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Disclose Ancillary Analyses

Mark subgroup, adjusted and sensitivity analyses clearly and distinguish planned from data-driven work.

Reporting move for Disclose Ancillary Analyses

Follow every denominator. The number invited, enrolled, assigned, exposed, followed and analysed can differ; a percentage is meaningful only when its denominator and condition are visible. Keep the heading **Disclose Ancillary Analyses** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A statistically adjusted result still depends on measured variables, their quality and the chosen model. The report should name the adjustment set and explain why it addresses the main design differences. For **disclose ancillary analyses**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **disclose ancillary analyses** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Report Adverse Events and Unintended Effects

Behavioural and public-health interventions can create burdens, stigma, displacement or unequal access as well as intended benefits.

Reporting move for Report Adverse Events and Unintended Effects

Pair every outcome estimate with its time point, precision and analysis population. Then label adjusted, subgroup and sensitivity analyses as planned or exploratory so flexibility remains visible. Keep the heading **Report Adverse Events and Unintended Effects** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

An unintended effect may be burden, stigma, displacement or unequal access rather than a clinical adverse event. TREND’s public-health purpose makes those consequences worth reporting explicitly. For **report adverse events and unintended effects**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **report adverse events and unintended effects** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Describe Intervention Fidelity

Compare intended delivery with actual exposure and explain how fidelity was assessed.

Reporting move for Describe Intervention Fidelity

End the section with a transfer sentence naming setting, staffing, participant mix and implementation conditions. This protects a useful local evaluation from becoming an unsupported universal promise. Keep the heading **Describe Intervention Fidelity** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

For a family reading a school programme evaluation, a responsible summary says which learners participated, how the comparison arose and what changed—not that every future learner will receive the same benefit. For **describe intervention fidelity**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **describe intervention fidelity** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Interpret Within the Nonrandomised Design

Discuss confounding, selection, measurement, missingness and generalisability before making causal claims.

Reporting move for Interpret Within the Nonrandomised Design

Reconstruct the evaluation as a timeline from recruitment through assignment, delivery, follow-up and analysis. A timeline exposes whether eligibility, intervention start and outcome measurement were aligned across conditions. Keep the heading **Interpret Within the Nonrandomised Design** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

Two schools may both appear in the intervention group because they volunteered early. If those schools also have more experienced staff, the allocation story belongs beside the outcome, not buried in a limitation paragraph. For **interpret within the nonrandomised design**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **interpret within the nonrandomised design** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Connect Findings to Programme Decisions

Explain what the evidence supports, what remains uncertain and which contextual conditions matter for transfer.

Reporting move for Connect Findings to Programme Decisions

Draw the units in nested boxes—community, organisation, class and person—and write the eligibility and assignment rule at each level. This prevents an individual-level conclusion from hiding a school-level allocation process. Keep the heading **Connect Findings to Programme Decisions** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A community campaign offered six sessions, but attendance ranged from zero to six. Reporting only assignment erases the difference between access and exposure; reporting only completers can create a selected comparison. For **connect findings to programme decisions**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **connect findings to programme decisions** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Did You Know? TREND Has 22 Items

The CDC describes TREND as a 22-item checklist for standardised reporting of nonrandomised controlled evaluations.

Reporting move for Did You Know? TREND Has 22 Items

Build an intervention exposure table with planned dose, offered dose, received dose and completed dose. Nonrandomised comparisons often become clearer when exposure is separated from mere assignment. Keep the heading **Did You Know? TREND Has 22 Items** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

If a policy starts in one district before another, calendar time may bring other changes. The report should show dates, concurrent programmes and measurement schedules before attributing every difference to the intervention. For **did you know? trend has 22 items**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **did you know? trend has 22 items** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Keep TREND Separate From CONSORT

TREND addresses nonrandomised evaluations, while CONSORT is designed for randomised trials.

Reporting move for Keep TREND Separate From CONSORT

Copy the baseline characteristics that matter causally, then explain why each might affect both participation or programme choice and the outcome. Do not turn the table into a ritual significance test. Keep the heading **Keep TREND Separate From CONSORT** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A flow diagram can reveal that participants with missing follow-up were more common in one condition. The next question is why they left and how the analysis handled information that may depend on outcome risk. For **keep trend separate from consort**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **keep trend separate from consort** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Keep TREND Separate From ROBINS-I

TREND supports complete reporting; ROBINS-I is a risk-of-bias tool for nonrandomised intervention results.

Reporting move for Keep TREND Separate From ROBINS-I

Follow every denominator. The number invited, enrolled, assigned, exposed, followed and analysed can differ; a percentage is meaningful only when its denominator and condition are visible. Keep the heading **Keep TREND Separate From ROBINS-I** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

A statistically adjusted result still depends on measured variables, their quality and the chosen model. The report should name the adjustment set and explain why it addresses the main design differences. For **keep trend separate from robins-i**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **keep trend separate from robins-i** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Help Students Explain the Counterfactual

Ask what would likely have happened to the intervention group without the programme and why the comparison group is informative.

Reporting move for Help Students Explain the Counterfactual

Pair every outcome estimate with its time point, precision and analysis population. Then label adjusted, subgroup and sensitivity analyses as planned or exploratory so flexibility remains visible. Keep the heading **Help Students Explain the Counterfactual** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

An unintended effect may be burden, stigma, displacement or unequal access rather than a clinical adverse event. TREND’s public-health purpose makes those consequences worth reporting explicitly. For **help students explain the counterfactual**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **help students explain the counterfactual** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Help Parents Read School Programme Claims

Look for group allocation, baseline differences, exposure, outcomes and limitations before treating an evaluation as a promise for every learner.

Reporting move for Help Parents Read School Programme Claims

End the section with a transfer sentence naming setting, staffing, participant mix and implementation conditions. This protects a useful local evaluation from becoming an unsupported universal promise. Keep the heading **Help Parents Read School Programme Claims** attached to the unit of allocation, the relevant dates and the group-specific evidence.

Concrete evaluation example

For a family reading a school programme evaluation, a responsible summary says which learners participated, how the comparison arose and what changed—not that every future learner will receive the same benefit. For **help parents read school programme claims**, translate the example back to the report by marking the page, table, figure or supplement where the necessary information appears.

Interpretation boundary

TREND makes **help parents read school programme claims** transparent; it does not create randomisation or eliminate confounding. After checking reporting completeness, apply design-appropriate appraisal and write what the comparison supports, which alternative explanations remain and what a new site would need to verify before transfer.

---

Frequently Asked Questions

What does TREND stand for?

Transparent Reporting of Evaluations with Nonrandomized Designs.

How many checklist items does TREND have?

The CDC describes a 22-item checklist.

Who developed TREND?

The CDC HIV Prevention Research Synthesis Project collaborated with HIV researchers and journal editors.

When was TREND first published?

The statement was first published in March 2004.

Does TREND prove an intervention works?

No. It improves reporting and does not remove confounding, selection bias or other design limitations.

Why report allocation if it was not random?

The rule or process explains why groups may differ before the intervention.

Why is participant flow important?

It reveals who was screened, enrolled, exposed, followed and analysed in each condition.

How is TREND different from CONSORT?

TREND is for nonrandomised evaluations; CONSORT is for randomised trials.

How is TREND different from ROBINS-I?

TREND is reporting guidance; ROBINS-I is a risk-of-bias tool.

Can school programme evaluations use TREND ideas?

Yes, when the design fits, especially for transparent allocation, exposure, flow and limitation reporting.

---

Useful Next Reading

• Read the current CDC TREND overview.
• Use the official 22-item TREND checklist.
• Describe intervention materials, dose and fidelity with TIDieR.
• Appraise a non-randomised result with ROBINS-I.
• Practise critical reading of claims, evidence and bias.

Closing thought: TREND turns a programme narrative into an inspectable evaluation by keeping allocation, delivery, participant flow, analysis and limitation in the same evidence chain.