Sample work

See what our feedback actually looks like.

One illustrative example per stage of the systematic review lifecycle — including the actual artifacts, like flow diagrams and forest plots, not just written notes.

These examples are illustrative of our typical feedback style and are not excerpts from real client projects, which remain confidential under our NDA.

[1] Protocol & Registration

PICO framing feedback

Protocol consulting excerpt

Feedback tightening a research question into a searchable protocol structure:

Original question: “What is the evidence on exercise for depression?”

Reframed as PICO: In adults with major depressive disorder (P), does structured aerobic exercise (I), compared to standard pharmacological treatment alone (C), reduce depressive symptom severity at 12 weeks (O)?

Specific enough to build a reproducible search strategy and defensible eligibility criteria, and to register cleanly on PROSPERO.

[2] Search Strategy

Search strategy peer review note

PRESS-style review excerpt

A note from a search strategy peer review, before the search was run:

Flag: MeSH term “Diabetes Mellitus” used without the exploded subheadings, missing Type 2-specific subject headings in 2 of 4 databases.

Fix: Added exploded MeSH tree plus free-text synonyms (“T2DM”, “adult-onset diabetes”). Recall increased from an estimated 60% to 90%+ coverage on a test set of 12 known-relevant studies.

[3] Screening & Extraction

Screening conflict resolution note

Dual-review consulting excerpt

A documented conflict-resolution note between two independent screeners, and the flow diagram it feeds into:

Disagreement: Reviewer A included, Reviewer B excluded a study on setting grounds (inpatient vs. outpatient population).

Resolution: Re-read against the protocol’s eligibility criteria — study population was mixed-setting, meeting inclusion criteria. Included, with the ambiguity noted in the PRISMA flow diagram.

Resulting PRISMA flow diagramRecords identifiedn = 2,314After deduplicationn = 1,802Title/abstract screenedn = 1,802Full-text assessedn = 96Studies includedn = 34Excluded (full-text)n = 62
[4] Risk of Bias & GRADE

Risk-of-bias reviewer note

Methodology consulting excerpt

A consultant note flagging a risk-of-bias tool mismatch before it reached peer review:

Flag: Included studies are a mix of RCTs and cohort designs, but only Cochrane RoB 2 has been applied.

Recommendation: Apply ROBINS-I to the 6 non-randomized studies (IDs 4, 7, 12, 15, 19, 22) instead. RoB 2 assumes randomization; using it on cohort studies will draw a methods-reviewer objection.

Est. time to resolve: ~2 hrs per study, reviewer-ready appraisal table included in the next round.

[5] Meta-Analysis

Heterogeneity interpretation note

Statistical consulting excerpt

Consulting notes interpreting a heterogeneity result before finalizing the pooling model, alongside the resulting forest plot:

Result: I² = 78% across 14 included studies — substantial heterogeneity under a fixed-effects assumption.

Recommendation: Switch to a random-effects model, and run a pre-specified subgroup analysis by intervention duration (<8 weeks vs. ≥8 weeks) to explore the source of variation.

Subgroup analysis explained the majority of the heterogeneity — reported transparently rather than masked by pooling.

Resulting forest plotStudy A (2021)Study B (2022)Study C (2020)Study D (2023)Study E (2021)Pooled estimateFavors controlFavors intervention
[6] Manuscript & Review Editing

PRISMA compliance edit

Line edit

The kind of specific, PRISMA-checked edit every manuscript client receives:

The findings shows findings show that participant retention were was significantly higher across the intervention cohort than in the control group, which recorded a 34% dropout rate by week six.

Reviewed for PRISMA 2020 compliance, clarity & APA 7 · 12 edits · 0 rewrites

Want feedback like this on your own review?

Get a Quote →