返回提示词目录第 1468 页

公共卫生信息可信度审查

系统检查公共卫生说法的证据质量、风险表达、偏差与不确定性,输出可核验的分析结论。

完整提示词

Main objective:

> Audit a public-health claim for source quality, cherry-picking, causal overreach, denominator errors, fear framing and contradiction with stronger evidence.

## 1. POPULATION CONTEXT

Define population, geography, time period, baseline risk, exposure/intervention, comparator, outcome, surveillance system, source date and policy/clinical decision being informed.

## 2. EVIDENCE & SURVEILLANCE

Prioritize current public-health agencies, WHO, national surveillance, high-quality systematic reviews and appropriate epidemiologic studies. Distinguish preliminary surveillance signals from confirmed causal evidence.

## 3. RISK COMMUNICATION

Always show where possible:
- baseline / absolute risk
- relative risk or effect
- timeframe
- denominator
- uncertainty interval
- population to which the estimate applies
- whether the number describes infection, disease, hospitalization, death, screening detection or another outcome

Avoid percentage framing without denominators.

## 4. SPECIALIZED TEST

For **Public Health Misinformation Audit**:
- Audit a public-health claim for source quality, cherry-picking, causal overreach, denominator errors, fear framing and contradiction with stronger evidence.
- test source recency and geographic applicability
- identify selection, measurement and reporting bias
- distinguish individual risk from population effect
- include benefit, harm and implementation burden
- assess differential impact on vulnerable or underserved groups
- state what data would change the conclusion

## 5. MISINFORMATION DEFENSE

Check for:
- cherry-picked time windows
- numerator without denominator
- ecological fallacy
- association presented as causation
- relative risk without absolute risk
- surveillance artifact treated as biological change
- preprint / social post treated as settled evidence
- subgroup claim from underpowered analysis
- false balance between unequal evidence bases
- outdated guidance presented as current

## 6. REQUIRED MATRICES

### Population Evidence Matrix
| Source | Population | Period | Exposure/intervention | Outcome | Absolute effect | Relative effect | Bias | Certainty |
|---|---|---|---|---|---|---|---|---|

### Equity Matrix
| Group | Exposure / access | Baseline risk | Benefit | Harm / burden | Data gap |
|---|---|---|---|---|---|

### Communication Matrix
| Claim | Correct denominator | Timeframe | Plain-language meaning | Common misread |
|---|---|---|---|---|

## 7. REQUIRED OUTPUT

1. Population and timeframe.
2. Best current evidence.
3. Absolute and relative effects.
4. Benefit-harm balance.
5. Bias and uncertainty.
6. Equity impact.
7. Misinterpretation risks.
8. Calibrated public-health conclusion.

End with **Public Health Evidence Integrity Check** confirming current source date, denominator, timeframe, population and uncertainty are explicit.

This supports public-health education and analysis, not individual medical diagnosis or treatment.

分类与标签

思维提升、信息核查、公共卫生、证据分析、风险沟通

前往 AI 图片创作