> ## Documentation Index
> Fetch the complete documentation index at: https://docs-evisyn.grabmywork.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Search Strategy

> AI-powered PICO decomposition and database-specific query generation

The search strategy is the foundation of your systematic review. EviSyn's AI analyzes your research question and generates structured, database-specific search queries — a task that traditionally takes 20-40 hours of expert time.

## Step 1: Configure Your Strategy

Navigate to **Search Strategy** from your project dashboard.

<img src="https://mintcdn.com/ravian/YwBZFdlBbtQ_-NDF/images/search-strategy.png?fit=max&auto=format&n=YwBZFdlBbtQ_-NDF&q=85&s=2c216caae3cc793e8df6a93edfab91eb" alt="Search strategy configuration" width="2278" height="1632" data-path="images/search-strategy.png" />

### Settings

| Setting                | Options                                  | Description                                                                 |
| ---------------------- | ---------------------------------------- | --------------------------------------------------------------------------- |
| **Framework**          | PICO (Quantitative) / PICo (Qualitative) | PICO for intervention studies, PICo for qualitative research                |
| **Scope**              | Broad / Narrow                           | Broad captures more results with higher sensitivity; Narrow is more precise |
| **Target Databases**   | PubMed, EMBASE, Scopus, Cochrane         | Select which databases to generate queries for                              |
| **Additional Context** | Free text                                | Optional: add inclusion/exclusion criteria or study type filters            |

<Info>
  **PubMed** is available on all plans. EMBASE, Scopus, and Cochrane require API keys configured in [Settings](/user-guide/api-keys).
</Info>

### Previous Strategies

You can generate multiple strategies and compare them. Each strategy is saved with its date, database, and status (draft or executed).

Click **Generate PICO Strategy (AI)** to run the AI decomposition.

## Step 2: Review PICO Decomposition

The AI breaks your research question into structured PICO components, each with suggested search terms.

<img src="https://mintcdn.com/ravian/YwBZFdlBbtQ_-NDF/images/pico-decomposition.png?fit=max&auto=format&n=YwBZFdlBbtQ_-NDF&q=85&s=11e9fbdeb227dcd974377593e0f97f1e" alt="PICO decomposition results" width="2330" height="1182" data-path="images/pico-decomposition.png" />

### What You Get

<CardGroup cols={2}>
  <Card title="P — Population" icon="users" color="#FFF3E0">
    Who is being studied? Includes MeSH terms and free-text synonyms.

    *Example: "Patients planned for lung cancer surgery" → lung cancer, lung neoplasms, NSCLC, thoracic surgery*
  </Card>

  <Card title="I — Intervention" icon="syringe" color="#E8F5E9">
    What treatment or action is being applied?

    *Example: "Prehabilitation" → preoperative exercise, preoperative rehabilitation, preoperative conditioning*
  </Card>

  <Card title="C — Comparator" icon="scale-balanced" color="#E3F2FD">
    What is it compared against?

    *Example: "Standard of care" → usual care, no prehabilitation*
  </Card>

  <Card title="O — Outcome" icon="chart-line" color="#F3E5F5">
    What is being measured?

    *Example: "Perioperative outcomes" → postoperative complications, length of stay, mortality*
  </Card>
</CardGroup>

The AI also identifies:

* **Study Design** — RCTs, controlled clinical trials, cohort studies
* **Evidence Type** — Primary research, systematic reviews, meta-analyses

<Tip>
  You can edit the PICO terms before generating queries. Add or remove synonyms to adjust sensitivity.
</Tip>

## Step 3: Review & Edit Queries

The AI generates database-specific boolean queries for each PICO element and a combined query.

<img src="https://mintcdn.com/ravian/YwBZFdlBbtQ_-NDF/images/search-queries.png?fit=max&auto=format&n=YwBZFdlBbtQ_-NDF&q=85&s=3a521bbb6dcc7cb24aa41f88b4cd75fc" alt="Database-specific search queries" width="1794" height="1542" data-path="images/search-queries.png" />

### Per-Element Queries

Each PICO element has its own query panel showing:

* **Boolean query** with MeSH terms (for PubMed) or free-text terms
* **Get Count** button — preview how many results each element returns
* Editable text — modify queries before executing

### Combined Query

At the bottom, the **Combined Query** merges all elements using `AND` logic:

```
(Population terms) AND (Intervention terms) AND (Comparator terms) AND (Outcome terms)
```

You can:

* Click **Get Count** to preview total hits
* Click **Auto-Combine** to regenerate the combined query from current elements
* Edit the combined query directly

### Execute the Search

Click **Execute PUBMED Search** to run the query and import results directly into your project.

<Note>
  The hit count badge (e.g., "25 hits") shows how many papers matched. Papers are imported with full metadata: title, authors, journal, year, DOI, and abstract.
</Note>

## What's Next?

After executing your search (or uploading files manually), proceed to [Import Files](/user-guide/importing-papers) to review your imported papers.