| 当前可读状态与来源 | 打开此版本 → Combines comparable effect estimates across studies while accounting for sampling uncertainty and, where modeled, heterogeneity. 资料卡来源已关联;不等于专家适用性核验。 来源记录日期:2026-10-06 来源包 SHA-256:26634a4dabae17defbd1b24eb220be86a829dded1d20b5c86720d34d32e856fc |
|---|
| 适用问题 | - synthesizing comparable study estimates
- quantifying heterogeneity
|
|---|
| 输入 | - 可比研究的效应量、标准误或可换算的统计量
- 纳入规则、偏倚评估及合并模型
|
|---|
| 输出 | - 合并效应及不确定性、研究权重与异质性
- 敏感性结果;统计合并不修复原研究偏倚
|
|---|
| 数据条件 | - eligibility criteria
- effect estimates and uncertainty
- risk-of-bias assessment
|
|---|
| 假设 | - effects are meaningfully combinable under chosen model
- effect measures are compatible
- bias/missing-results risks considered
|
|---|
| 识别假设 | 此版本未提供 |
|---|
| 不能支持 | - valid synthesis of systematically biased evidence
- causal certainty beyond source-study designs
|
|---|
| 诊断 | - heterogeneity
- influence/sensitivity
- risk of bias
- missing-results/publication-bias assessment
|
|---|
| 常见失败 | - pooling incomparable studies
- ignoring bias
- overinterpreting pooled effect under extreme heterogeneity
|
|---|
| 替代方法 | - narrative synthesis
- network meta-analysis
- IPD meta-analysis
|
|---|
| 软件 | |
|---|
| 参考来源与核对范围 | - Cochrane Handbook Chapter 10: Analysing data and undertaking meta-analyses · 来源 ↗
|
|---|