返回提示词目录第 1440 页

循证实践证据表构建

按指定证据等级整理研究来源,提取设计、样本和结果,评定质量并形成综合结论与实践建议。

完整提示词

<context>
Evidence-based practice projects in nursing and allied health (practice change proposals, DNP and capstone projects, unit-level quality work) are built on an evidence table: one row per source with its design, sample, findings, and a level and quality rating on the hierarchy the organisation uses. In the Johns Hopkins model, levels run from I (experimental studies such as RCTs, and systematic reviews of RCTs) through II (quasi-experimental studies, and systematic reviews of quasi-experimental studies alone or mixed with RCTs), III (non-experimental and qualitative studies, and systematic reviews that include them), IV (clinical practice guidelines and consensus statements from recognised expert bodies) to V (literature reviews, quality improvement and programme evaluations, case reports, and expert opinion based on experience), with quality rated A (high), B (good) or C (low or major flaws) using the model's appraisal tools. Melnyk and Fineout-Overholt use seven levels: I systematic reviews or meta-analyses of RCTs, II single well-designed RCTs, III controlled trials without randomisation, IV case-control and cohort studies, V systematic reviews of descriptive and qualitative studies, VI single descriptive or qualitative studies, VII expert opinion. Common mistakes: rating a source by what it calls itself rather than its design, rating level and calling it quality, giving a guideline a research level, rating from the abstract, and a synthesis that counts sources instead of weighing them. Hierarchies are updated between editions, so the user's programme version governs.
</context>

<task>
Build an evidence table using the {{hierarchy}} hierarchy for this question:
<question>
{{question}}
</question>
{{#template}}
Required columns:
{{template}}
{{/template}}
<sources>
{{sources}}
</sources>

1. Restate the question in PICO form and note anything vague that would affect inclusion.
2. For each source, extract: citation as given, purpose, design (from the methods, not the label), setting and sample, intervention or exposure, outcomes and measures, key findings with numbers as reported, and limitations.
3. Assign the level on the chosen hierarchy with a one-line reason tied to the design. For "other", ask for the scheme's levels or use a clearly labelled generic design-based ranking.
4. Assign quality using the hierarchy's quality criteria (for Johns Hopkins A, B or C), giving the main reasons. Mark ratings made on limited text as provisional.
5. Summarise the body of evidence: counts by level and quality, consistency of findings, and an overall strength judgement in the hierarchy's own terms (for Johns Hopkins, along the lines of strong and compelling with consistent results, good and consistent, good but conflicting, or little or no evidence).
6. Synthesise by outcome, weighing higher-level and higher-quality sources more, and noting where findings disagree.
7. State a practice recommendation at the strength the evidence allows (for example consider a pilot, change practice, do not change practice, more evidence needed), with organisational fit and feasibility to check.
</task>

<constraints>
- Extract only from the text given; copy numbers exactly. Never fill a cell from memory; write "not reported".
- Rate design from the methods. If a source's label and methods disagree, rate the methods and note it.
- Rating quality needs the methods; a rating from an abstract alone is provisional and must be labelled so.
- Say that ratings should be checked against the hierarchy edition and appraisal tools the programme requires, and ideally done independently by two people.
- The recommendation informs a practice change process; it does not replace organisational approval, local policy or clinical judgement.
</constraints>

<output_format>
## Question
PICO and any notes.
## Evidence table
Table with the required columns, or: source | design | sample and setting | intervention | outcomes | key findings | limitations | level | quality.
## Level and quality summary
Counts and a one-line overall rating.
## Synthesis
Short paragraphs by outcome.
## Practice recommendation
Two or three sentences.
## Gaps and next steps
Bullets.
</output_format>

可替换内容

  • {{hierarchy}}
  • {{question}}
  • {{#template}}
  • {{template}}
  • {{/template}}
  • {{sources}}

分类与标签

学习教育、循证实践、文献综述、证据分级、研究评价

前往 AI 图片创作