literature-review-planner

文献综述规划

适用于学术和专业研究场景的文献综述规划、实施与综合的全面框架。

综述类型

类型	目的	范围	方法严谨度	适用场景
叙述性综述	广泛概述某一主题	范围广、灵活	低至中	背景章节、引言
系统综述	回答具体的研究问题	范围窄、预先设定	高	基于证据的决策、临床实践
范围综述	梳理某一主题的现有证据	范围广、结构化	中	新兴领域、识别研究空白
Meta-Analysis	定量综合研究结果	范围窄、统计分析	最高	合并效应量、评估治疗效果
快速综述	及时综合证据	聚焦、简化	中	政策决策、时间受限的场景
伞式综述	综述现有综述	仅限综述	高	综合整体证据
整合性综述	综合多种研究方法	范围广、混合方法	中	综合定性和定量研究

选择合适的综述类型

Do you need to answer a specific, focused question?
  YES --> Is quantitative synthesis of effect sizes needed?
    YES --> Meta-Analysis
    NO  --> Systematic Review
  NO --> Do you need to map the breadth of evidence?
    YES --> Is the field well-established?
      YES --> Umbrella Review (review of reviews)
      NO  --> Scoping Review
    NO --> Do you need to combine qualitative and quantitative?
      YES --> Integrative Review
      NO --> Is time constrained (< 3 months)?
        YES --> Rapid Review
        NO  --> Narrative Review

制定检索策略

PICO/PEO 框架

使用结构化框架来定义研究问题：

框架	要素	描述	示例
PICO	研究人群	研究对象是谁	患有 2 型糖尿病的成年人
	干预措施	采用何种治疗／暴露	远程医疗咨询
	对照措施	干预措施的替代方案	面对面咨询
	结局	测量什么	HbA1c 水平、患者满意度
PEO	研究人群	研究对象是谁	软件工程团队
	暴露	关注的现象	采用敏捷方法
	结局	测量什么	生产力、代码质量

数据库选择

数据库	覆盖范围	适用场景
PubMed/MEDLINE	生物医学、生命科学	临床、医学、健康研究
Scopus	多学科、覆盖面最广	跨学科综述
Web of Science	多学科、引文数据	引文分析、影响力追踪
IEEE Xplore	工程、计算机科学	技术与计算研究
PsycINFO	心理学、行为科学	心理健康、认知研究
ERIC	教育	教学、学习、教育政策
CINAHL	护理、相关健康领域	护理及卫生专业研究
Cochrane Library	系统综述、临床试验	临床干预证据
Google Scholar	覆盖面广、灰色文献	补充检索、滚雪球检索
预印本服务器	arXiv、bioRxiv、SSRN	前沿的未发表研究

关键词与布尔检索策略

BUILDING A SEARCH STRING:

Step 1: Identify key concepts from PICO/PEO
  Concept 1: "telemedicine" OR "telehealth" OR "remote consultation" OR "virtual care"
  Concept 2: "diabetes" OR "type 2 diabetes" OR "T2DM" OR "diabetes mellitus"
  Concept 3: "glycemic control" OR "HbA1c" OR "blood glucose" OR "patient outcomes"

Step 2: Combine with Boolean operators
  (Concept 1) AND (Concept 2) AND (Concept 3)

Step 3: Apply filters
  - Date range: 2015-2025
  - Language: English
  - Study type: RCT, cohort, systematic review
  - Peer-reviewed only

ADVANCED OPERATORS:
  "exact phrase"      - Exact match
  *                   - Truncation (therap* = therapy, therapies, therapeutic)
  MeSH terms          - Controlled vocabulary (PubMed)
  NEAR/3              - Proximity (terms within 3 words)
  ti,ab               - Title and abstract search

检索记录模板

SEARCH LOG:

Database: [Name]
Date Searched: [Date]
Search String: [Full query]
Filters Applied: [Date, language, study type]
Results Retrieved: [Count]
Results After Deduplication: [Count]
Notes: [Any issues, modifications needed]

PRISMA 流程图

IDENTIFICATION
  Records identified through database searching: n = ___
  Records identified through other sources: n = ___
  |
  v
  Records after duplicates removed: n = ___
  |
SCREENING
  v
  Records screened (title/abstract): n = ___
  Records excluded: n = ___
  |
  v
  Full-text articles assessed for eligibility: n = ___
  Full-text articles excluded (with reasons): n = ___
    - Reason 1: n = ___
    - Reason 2: n = ___
    - Reason 3: n = ___
  |
INCLUDED
  v
  Studies included in qualitative synthesis: n = ___
  Studies included in quantitative synthesis (meta-analysis): n = ___

纳入／排除标准

标准	纳入	排除
研究人群	[定义目标人群]	[定义排除的人群]
干预措施／暴露	[定义相关干预措施]	[定义排除的干预措施]
结局	[定义相关结局]	[与研究无关的结局]
研究设计	[可接受的研究类型]	[排除的研究类型]
日期范围	[起始年份] 至 [结束年份]	日期范围以外
语言	[可接受的语言]	其他语言
出版类型	同行评审期刊	社论、信件、会议摘要

来源评估

按研究设计划分的关键评价工具

研究设计	评价工具	关键领域
RCTs	Cochrane 偏倚风险评估工具（RoB 2）	随机化、盲法、失访、报告
队列研究	Newcastle-Ottawa 量表（NOS）	选择、可比性、结局评估
病例对照研究	Newcastle-Ottawa 量表（NOS）	选择、可比性、暴露评估
定性研究	CASP 定性研究核查表	目标、方法学、招募、数据、分析、伦理
横断面研究	JBI 批判性评价工具	纳入、测量、混杂因素、分析
诊断研究	QUADAS-2	患者选择、指标检验、参考标准、流程
混合方法研究	MMAT	定性、定量、混合方法标准

来源质量评估框架

QUALITY SCORING (rate each 1-5):

RELEVANCE:
  - Directly addresses research question? ___
  - Population matches target? ___
  - Outcomes align with review objectives? ___

METHODOLOGICAL RIGOR:
  - Study design appropriate? ___
  - Sample size adequate? ___
  - Bias minimized? ___
  - Statistical analysis appropriate? ___

CREDIBILITY:
  - Published in peer-reviewed journal? ___
  - Authors have relevant expertise? ___
  - Funding sources declared? ___
  - Conflicts of interest addressed? ___

RECENCY:
  - Published within target date range? ___
  - Findings still applicable? ___
  - Not superseded by newer evidence? ___

TOTAL SCORE: ___ / 60
  High quality: 48-60
  Medium quality: 36-47
  Low quality: < 36

证据层级

EVIDENCE PYRAMID (highest to lowest):

Level 1: Systematic Reviews & Meta-Analyses
Level 2: Randomized Controlled Trials (RCTs)
Level 3: Cohort Studies (prospective)
Level 4: Case-Control Studies
Level 5: Cross-Sectional Studies / Case Series
Level 6: Expert Opinion / Editorials
Level 7: Anecdotal / Narrative Reports

引文管理

工作流

CITATION MANAGEMENT PROCESS:

1. COLLECT
   - Export references from databases (RIS, BibTeX, EndNote XML)
   - Import into reference manager (Zotero, Mendeley, EndNote)
   - Attach PDFs where available

2. ORGANIZE
   - Create folder structure mirroring review themes
   - Tag with inclusion/exclusion status
   - Tag with quality rating
   - Add notes and annotations

3. DEDUPLICATE
   - Run automatic deduplication
   - Manual review of near-duplicates
   - Document count removed

4. SCREEN
   - Title/abstract screening (tag: include/exclude/maybe)
   - Full-text screening (tag: include/exclude with reason)
   - Track screening decisions

5. EXTRACT
   - Populate data extraction form
   - Link to source reference
   - Note discrepancies

数据提取模板

EXTRACTION FORM:

Study ID: ___
Authors: ___
Year: ___
Title: ___
Journal: ___
Study Design: ___
Country/Setting: ___

Population:
  - Sample size: ___
  - Demographics: ___
  - Inclusion criteria: ___

Intervention/Exposure: ___
Comparison/Control: ___

Outcomes:
  - Primary: ___
  - Secondary: ___
  - Measurement tools: ___

Key Findings: ___
Effect Size (if applicable): ___
Confidence Interval: ___
Quality Rating: ___
Reviewer Notes: ___

综合框架

主题综合

THEMATIC SYNTHESIS STEPS:

1. CODE: Read included studies and assign descriptive codes
2. ORGANIZE: Group related codes into descriptive themes
3. DEVELOP: Generate analytical themes that go beyond the primary studies
4. MAP: Create a thematic map showing relationships between themes
5. WRITE: Narrate findings organized by analytical themes

THEMATIC MAP STRUCTURE:
  Overarching Theme
  |-- Sub-theme 1
  |   |-- Code A (Studies 1, 3, 7)
  |   |-- Code B (Studies 2, 5)
  |-- Sub-theme 2
  |   |-- Code C (Studies 1, 4, 6)
  |   |-- Code D (Studies 3, 8)

时间顺序综合

最适合展示对某个主题的理解如何随时间演变。

CHRONOLOGICAL STRUCTURE:

Era 1 (e.g., 2000-2010): Foundational Work
  - Key studies and their contributions
  - Prevailing theories and methods

Era 2 (e.g., 2010-2018): Methodological Advances
  - New approaches introduced
  - Challenges to earlier findings

Era 3 (e.g., 2018-Present): Current State
  - Latest findings and debates
  - Emerging directions

方法学综合

按方法学对研究分组，以比较不同方法如何得出不同见解。

方法学	研究	关键发现	优势	局限性
RCTs	[list]	[summary]	因果推断	可推广性
定性研究	[list]	[summary]	丰富的背景信息	主观性
混合方法研究	[list]	[summary]	全面性	复杂性
观察性研究	[list]	[summary]	真实世界有效性	混杂因素

研究空白识别

空白分析框架

GAP CATEGORIES:

KNOWLEDGE GAPS:
  - What questions remain unanswered?
  - Where do findings conflict?
  - What populations are understudied?

METHODOLOGICAL GAPS:
  - What study designs are missing?
  - Are sample sizes consistently too small?
  - Are measurement tools validated?

CONTEXTUAL GAPS:
  - What geographic regions are underrepresented?
  - What settings haven't been studied?
  - Are there temporal gaps in the literature?

PRACTICAL GAPS:
  - What interventions haven't been tested?
  - Where does evidence fail to translate to practice?
  - What implementation barriers are unaddressed?

空白记录模板

GAP: [Brief description]
EVIDENCE: [What the current literature shows / doesn't show]
SIGNIFICANCE: [Why this gap matters]
SUGGESTED RESEARCH: [What future studies could address this]
PRIORITY: [High / Medium / Low]

写作结构

文献综述章节

STRUCTURE:

1. INTRODUCTION (10-15% of word count)
   - Context and importance of the topic
   - Scope and objectives of the review
   - Research question(s)
   - Brief overview of structure

2. METHODOLOGY (15-20% for systematic; shorter for narrative)
   - Search strategy and databases
   - Inclusion/exclusion criteria
   - Screening process (PRISMA for systematic)
   - Quality assessment approach
   - Data extraction method
   - Synthesis approach

3. FINDINGS / RESULTS (40-50%)
   - Organized by themes, chronology, or methodology
   - Summary tables of included studies
   - Critical analysis (not just description)
   - Comparison and contrast across studies
   - Quality assessment results

4. DISCUSSION (15-20%)
   - Synthesis of key findings
   - Comparison with existing reviews
   - Implications for theory and practice
   - Strengths and limitations of the review

5. GAPS AND FUTURE DIRECTIONS (5-10%)
   - Identified gaps in knowledge
   - Recommended research priorities
   - Methodological recommendations

6. CONCLUSION (5%)
   - Summary of main findings
   - Answer to research question
   - Key implications

常见问题

问题	描述	预防措施
挑选性引用	仅选择支持某个假设的研究	预注册方案，遵循 PRISMA
叙述偏差	只描述研究，不进行批判性分析	使用评价工具，跨研究比较
范围蔓延	将重点扩展到原始问题之外	遵循预先设定的纳入标准
近因偏差	过度重视近期研究	纳入完整的日期范围，按质量赋予权重
发表偏差	遗漏灰色文献和无效结果	检索预印本、学位论文和试验注册库
检索不足	数据库过少或检索词范围过窄	至少检索 3 个数据库，迭代优化检索策略
综合不佳	罗列研究，而非整合研究结果	使用综合框架，识别模式
缺少方案	未预注册综述方案	开始前在 PROSPERO 或 OSF 注册

综述方案模板

PROTOCOL:

Title: [Review title]
Registration: [PROSPERO/OSF ID]
Authors: [Names and roles]
Date: [Protocol date]

Background: [Why this review is needed]
Objectives: [What the review aims to achieve]
Research Question: [PICO/PEO formatted question]

Eligibility Criteria:
  Inclusion: [List]
  Exclusion: [List]

Information Sources: [Databases and other sources]
Search Strategy: [Full search string per database]

Study Selection:
  - Stage 1: Title/abstract screening (2 independent reviewers)
  - Stage 2: Full-text screening (2 independent reviewers)
  - Disagreement resolution: [Process]

Data Extraction: [What data will be extracted]
Quality Assessment: [Which tool(s) will be used]
Synthesis Method: [Narrative, thematic, meta-analysis]
Timeline: [Planned completion date]

另请参阅

• 数据科学
• 资助申请书构建器