Strata Academy

12-Week Systematic Review Timeline for Student Dissertations

Week-by-week plan from PICO and PROSPERO through search, screening, extraction, synthesis, and write-up for UK medical students

1. Before you start: scope and feasibility

Twelve weeks is tight but achievable for a focused intervention review with clear RCT evidence — not for an open-ended 'everything about diabetes' topic. Supervisors should agree on one primary outcome, eligible study designs, and whether meta-analysis is realistic.

Check PROSPERO for duplicate registrations. Run a scoping search in MEDLINE or PubMed and count recent hits — hundreds of RCTs may require narrowing; zero hits means rethink the PICO.

Assign roles if paired students: one leads search, one leads screening tool setup — but both must dual-screen a proportion of records.

• One primary outcome, pre-specified in PROSPERO
• Agree databases: minimum MEDLINE + Embase or CENTRAL equivalent
• Choose screening tool early (Rayyan, Covidence, SysRev)
• Book supervisor checkpoints at Weeks 2, 6, and 10

2. Weeks 1–2: Question, protocol, PROSPERO

Draft structured PICO, inclusion and exclusion criteria, and planned outcomes. Write a one-page protocol summary for supervisor sign-off.

Submit PROSPERO registration with search sources, study design eligibility, ROB tool (ROB 2 for RCTs), and planned synthesis (meta-analysis vs narrative).

While PROSPERO is under review, build search concepts and test keywords — do not wait idle.

1. Week 1: Finalise PICO, scoping search, supervisor agreement on feasibility
2. Week 1: Draft inclusion/exclusion criteria tied to screening decisions
3. Week 2: Submit PROSPERO; obtain CRD number when published
4. Week 2: Pilot search string in one database; refine MeSH and synonyms

3. Weeks 3–4: Full search and screening setup

Run final searches across agreed databases and registries. Export to screening software with deduplication documented.

Pilot screening on 50–100 records with both reviewers. Calculate inter-rater agreement (Cohen's kappa or percent agreement) and refine criteria if disagreement is high.

Begin title/abstract screening only after PROSPERO is registered or supervisor approves retrospective registration with documented rationale.

1. Week 3: Execute full search (MEDLINE, Embase/CENTRAL, registries); export and deduplicate
2. Week 3: Configure screening tool; import records; assign blinded reviewers
3. Week 4: Pilot dual screening; resolve conflicts; update criteria if needed
4. Week 4: Begin full title/abstract screening — target 50% completion by week end

4. Weeks 5–7: Screening completion and full-text review

Complete title/abstract screening with conflict resolution by a third reviewer or discussion. Document excluded studies at full-text stage with reasons — PRISMA 2020 expects this.

Obtain full texts; track failures (paywalls, missing appendices). Full-text screening is slower than students expect — budget two full weeks.

Update PRISMA flow counts as you go — reconstructing numbers at the end causes errors examiners spot immediately.

1. Week 5: Complete title/abstract screening; begin full-text retrieval
2. Week 6: Dual full-text screening; record exclusion reasons (wrong population, wrong design, wrong outcome)
3. Week 7: Finalise included study list; supervisor review of borderline papers
4. Week 7: Draft PRISMA 2020 flow diagram with current counts

5. Weeks 8–9: Data extraction and risk of bias

Build extraction form from protocol outcomes: study design, n, demographics, intervention, comparator, outcome data at time points, funding, conflicts.

Pilot extraction on two studies; reconcile forms before extracting the full set.

Apply ROB 2 (or ROBINS-I for non-randomised studies) with dual independent assessment where feasible. Summarise ROB in traffic-light plot.

1. Week 8: Pilot and complete data extraction; resolve discrepancies
2. Week 8: Enter numeric outcome data into synthesis spreadsheet or RevMan/R
3. Week 9: Complete ROB 2 assessments per included RCT
4. Week 9: Draft ROB summary figure and narrative for results section

6. Weeks 10–11: Synthesis, meta-analysis, and GRADE

Decide per outcome whether meta-analysis is appropriate (clinical homogeneity, same effect measure). Run random-effects models as default when in doubt.

Produce forest plots for pooled outcomes; funnel plots if ≥10 studies. Draft simplified Summary of Findings table with GRADE certainty per primary outcome.

Narrative synthesis outcomes that cannot be pooled — do not force meta-analysis on incompatible data.

1. Week 10: Run meta-analysis for pre-specified outcomes; explore heterogeneity (I², subgroup if protocol pre-specified)
2. Week 10: Sensitivity analyses (exclude high ROB studies, leave-one-out)
3. Week 11: Complete GRADE / SoF table for primary outcomes
4. Week 11: Draft results and discussion — clinical interpretation, limitations, protocol deviations

7. Week 12: Write-up, PRISMA, and submission buffer

Finalise methods (search dates, databases, screening process, ROB tools, software), results (PRISMA flow, study characteristics table, synthesis), and discussion (summary of evidence, limitations, implications).

Complete PRISMA 2020 checklist appendix. Cross-check PROSPERO protocol against final manuscript — document amendments.

Reserve three days minimum for formatting, references, figure resolution, and supervisor final read — not for new screening.

1. Days 1–2: Complete introduction, methods, results drafts
2. Days 3–4: Discussion, conclusions, abstract with PROSPERO ID
3. Days 5–6: PRISMA checklist, references, tables and figures
4. Days 7+: Buffer for supervisor feedback, submission portal, examiner formatting

Interactive version (quizzes, walkthroughs) loads when JavaScript is enabled.