Thematic analysis turns a stack of findings from medical studies into clear patterns that answer a tight review question. In a review, you do not interview participants; you work with the published findings and author interpretations. The goal is to code those findings, build themes that cut across papers, and write a synthesis that clinicians and researchers can use. This guide walks through a practical route that fits clinical topics, keeps a clean audit trail, and meets common reporting standards.
What Thematic Analysis Does In A Medical Review
In medicine, reviews often mix designs and settings. Thematic analysis helps by mapping shared ideas across that mix. You read the results and discussion sections of each paper, extract the relevant segments, label those segments with short codes, then group codes into themes. Themes do not copy a single study; they reflect the combined signal across studies.
Before you start coding, set a few ground rules. Decide whether you will code inductively from the data or with a prior lens linked to your question. Decide whether your codes will stick to semantic content or reach for latent meaning. Agree on how much structure your team will use for the codebook. The table below summarises common choices and how to pick among them.
| Decision | Options | Good Questions To Ask |
|---|---|---|
| Orientation | Inductive or deductive | Will prior theory steer coding, or will codes grow from the findings? |
| Meaning Level | Semantic or latent | Do you stay with what authors wrote, or read beneath the line? |
| Code Structure | Reflexive or codebook | Will codes evolve fluidly, or sit in a shared, fixed list? |
| Unit | Findings sections or quoted data | Will you code author interpretations, participant quotes, or both? |
| Review Scope | Scoping, narrative, or qualitative evidence review | How exhaustive is the search and appraisal, and why? |
| Appraisal Tool | CASP, JBI, or none | Will you rate study quality and use it in the synthesis? |
| Software | NVivo, ATLAS.ti, Taguette, or Excel | What keeps your team fast, transparent, and consistent? |
For step naming, many teams draw on Braun and Clarke’s thematic analysis, then adapt it to fit secondary data from published studies. For reporting, the PRISMA 2020 statement sets clear expectations for flow diagrams and items to include.
Doing Thematic Analysis For A Medical Literature Review
This step-by-step route keeps the work organised and reproducible while leaving room for insight.
1) Set A Tight Review Question
Define the patient group, the phenomenon of interest, and the setting. Keep language plain. Write down what counts as “in scope” and what does not. This small investment stops drift when you hit study number twenty.
2) Plan Search And Screening
Build a search plan with subject headings and free text. Record databases, dates, and limits. Pilot the plan, then run it. Use two screeners for a sample to align decisions, then finish the rest. Log reasons for exclusion so the PRISMA diagram is easy to fill later. If your topic combines trials and qualitative studies, tag the types early; it saves time during data extraction.
3) Extract The Right Data
Create a simple template to capture citation details, sample and setting, method, and the segments you will code. Pull text from the results and discussion where authors present findings, not methods or background. If a study reports both quotes and interpretations, keep them linked.
4) Get Familiar With The Material
Read a few studies end to end. Write quick memos: what stands out, what repeats, what jars with your expectations. These notes help later when themes need clear boundaries.
5) Code The Findings
Work line by line through extracted segments. Write short, active labels that capture the point of each segment. Stay close to the wording at first. Merge duplicate labels only when you are sure they mean the same thing. If you work as a team, meet early to compare a few pages and agree on naming rules.
6) Build Candidate Themes
Group related codes. Map links between groups. Name each group with a short phrase that describes its core idea. Check that each group draws on multiple studies, not just one. Where a group mixes different ideas, split it. Where two groups say the same thing, merge them.
7) Review Against The Data And Question
Read coded extracts inside each group. Ask: does the data fit the label; does the label answer the review question; is anything missing. Shuffle codes until each theme feels tight and distinct.
8) Define, Name, And Order Themes
Write a one-line essence for each theme. Draft a short paragraph per theme that states the idea, shows variation, and notes any limits. Order themes so the story flows for a clinical reader.
9) Appraise Confidence In The Synthesis
Note how study quality, sample breadth, coherence across studies, and data richness back or weaken each theme. A short table that links these points to each theme helps readers judge confidence.
10) Write And Report
Use theme labels as subheadings. Weave brief study examples to show the idea. Cite the paper when you quote a line. Include a flow diagram and the items listed in the PRISMA 2020 checklist. State limits and what readers should and should not do with the findings.
Thematic Analysis Steps For Medicine Reviews
The checklist below turns the route above into actions you can assign and track.
Scope And Protocol
- Write a short protocol that fixes the question, eligibility rules, and the plan for coding and theme building.
- Note team roles, meeting points, and how you will resolve differences.
Search And Select
- Run the search and export to a manager that removes duplicates cleanly.
- Screen titles and abstracts, then full texts. Keep a log of exclusions with plain reasons.
Data Handling And Coding
- Extract findings into a table or software project. Keep source links for quick checks.
- Set naming rules for codes. Example rules: use verbs; avoid study IDs in code names; write one idea per code.
- Code in batches and meet to compare. Update the list of codes after each meeting.
Theme Development
- Cluster codes with sticky notes or software queries. Draw quick maps of links between clusters.
- Test each cluster with fresh extracts from studies that were not in the first batch.
- Record theme changes in a log with date, change, and reason.
Write-Up And Checks
- Draft results with theme subheadings and tight study examples.
- Run a PRISMA check and a copy edit for tense, spelling, and table labels.
- Ask a clinical colleague to read the theme summaries for clarity.
When To Use Thematic Synthesis
Thematic synthesis is a close cousin to thematic analysis and suits reviews that draw on qualitative findings. It starts with coding, moves to descriptive themes, then builds analytic themes that answer the review question. The approach is outlined in the Cochrane Handbook chapter on qualitative evidence. For mixed methods reviews, you can place the themes next to quantitative results to show where the messages align or differ.
Appraisal, Reflexivity, And Team Habits
Appraisal helps clear judgements about confidence. Tools from JBI or CASP help you record issues with sampling, data collection, or analysis and how these might affect the synthesis. Do not turn appraisal into a score; describe the issues and their likely influence on each theme.
Reflexivity guards against drift. Write short notes about your clinical background, your stake in the topic, and how those views could tilt coding choices. Keep these notes with the audit trail.
Good team habits speed the work: fixed file names, tidy folders, short weekly meets, and a standing log of decisions. The table below lists what to save so you can track every change without heavy admin.
| Step | Evidence To Save | Example File Or Note |
|---|---|---|
| Protocol | Question, scope, rules | protocol_v1.docx; change log |
| Search | Strategies, dates, exports | search_strings.pdf; database_exports.zip |
| Screening | In/out decisions and reasons | screen_log.xlsx; PRISMA counts |
| Extraction | Segments with source links | extraction_sheet.xlsx; study_notes.md |
| Coding | Code list versions | codes_v1.csv, codes_v2.csv with dates |
| Theme Maps | Pictures of clusters and links | theme_map_2025-09-10.png |
| Decisions | What changed and why | audit_log.md with dates and initials |
| Write-Up | Drafts and peer comments | results_v1.docx; comments.pdf |
Common Pitfalls And Fast Fixes
Starting With Themes From A Different Topic
Preloading your head with someone else’s themes can box you in. Read two or three studies fresh, code them without old labels, then compare with prior work.
Letting One Big Study Dominate
A large trial process paper or a national cohort report can drown out smaller studies. Check whether each theme draws on more than one source and setting.
Over-Merging Codes
If you merge too early, subtle differences vanish. Keep near-duplicates apart until late. When you merge, write a one-line note that justifies the move.
Vague Theme Names
A theme name like “Barriers” says little. Add the object and angle, such as “Workload barriers to timely follow-up in primary care.”
Quotes Without Context
When you use a quote, tie it to the study ID and sample. One clear quote per point beats a string of lines.
Write-Up Template You Can Reuse
Abstract
State the question, sources, number of studies, and the top themes in two lines each.
Methods
Describe search dates and sources, eligibility rules, extraction, coding, theme building, appraisal, and how you judged confidence. Point to your protocol if available.
Results
Start with the PRISMA flow. Summarise study settings and samples. Present each theme with a crisp label, a short paragraph, and one striking example from a study.
Discussion
Tell the reader what the themes mean for practice or policy. Note gaps where more research would change action. Keep claims modest and tied to the data.
Reporting
Include the PRISMA diagram and checklist, and a short note on reflexivity and limits. Add a link to data or code lists if your journal allows it.
Mini Worked Example
Review question: How do adults with heart failure manage salt intake at home?
Extracted segments from two studies mention “reading labels,” “guessing portion size,” and “family meals set by others.” Codes: label reading; portion guessing; family control. Candidate theme: “Everyday food routines strain self-management.” The theme draws on different settings and links to the review question. A second theme, “Conflicting advice from clinics and leaflets,” grows from codes like “nurse said one thing,” and “leaflet gave a different limit.” With two themes named and defined, you can write short summaries and pick a clean quote from each study to show the idea.
Tools And Formats That Keep Work Flowing
Pick tools that match team size and budget. A solo reviewer can code in a tidy spreadsheet; a larger team may want NVivo or ATLAS.ti for shared projects and queries. Taguette is a free option with a gentle learning curve. For diagrams, simple slides or draw.io files do the job. Keep everything in a versioned folder with dates in file names.
For writes, plain text with markdown works well for notes and logs. Export tables to CSV so they can move between tools. Save one PDF per study with a standard file name: first author, year, short title.
Ethics And Credit
You are working with published material, yet treat quotes with care. Retain any anonymisation in the source. Credit each study when you use a quote or close paraphrase. If an author shares extra material, store it securely and note any terms for use.
Pre-Submission Checklist
- Clear review question and scope
- Search plan, logs, and exports saved
- Screening reasons logged for the PRISMA flow
- Extraction sheet with linked sources
- Code list with date stamps
- Theme maps and short definitions
- Appraisal summary and confidence notes
- Results written with theme labels and study examples
- PRISMA items attached and checked
With a tight plan, tidy records, and careful theme naming, thematic analysis inside a medical review reads cleanly and guides sound action. Keep coding notes tidy. Now.