How to Compare Systematic Review Consulting Providers

Choosing who supports a systematic review is a methodology decision, not just a vendor decision -- the wrong fit can cost you months of rework or a rejected protocol. Most providers describe themselves in similar terms: experienced, rigorous, PRISMA-compliant. The differences that actually matter show up in three specific places, and each one is checkable if you know what to ask.

Screening methodology: ask for the process, not the promise

Dual independent screening -- two reviewers assessing each study without seeing each other's decisions first -- is the methodological standard for a reason: it catches the errors a single reviewer makes alone, and it's what AMSTAR-2 and most journal editors expect to see. The question worth asking directly is not "do you do dual screening," since almost every provider will say yes. Ask instead: what tool or process do you use to track it, and what happens when the two reviewers disagree? A provider with a real process can describe their conflict-resolution step specifically -- discussion, a third reviewer, documented reasoning -- without hesitating. A vague answer here is a genuine signal, not a minor gap.

The same logic applies to data extraction. Ask whether extraction is single-reviewer with spot-checking, or fully dual and independent like screening. Both are defensible choices depending on your review's risk tolerance and budget, but you want to know which one you're actually getting, and you want it stated in whatever scope document or contract you sign -- not just described verbally in a sales conversation.

Reviewer expertise: match to your specific topic, not a general credential

"PhD-level reviewers" is a claim that tells you almost nothing on its own. A methodologist with a strong general systematic review background can still be a poor fit for a review requiring deep familiarity with, say, diagnostic accuracy studies or a specific clinical subspecialty -- the statistical and appraisal considerations genuinely differ by review type. Ask specifically who will work on your review, what their relevant background is, and whether that person (not just the company as a whole) has handled your review type or topic area before.

This matters most for risk-of-bias and quality appraisal, where subject-matter judgment genuinely affects the outcome -- assessing confounding in a non-randomized study, for instance, requires understanding which confounders are actually plausible for your specific intervention and population, not just knowing how to fill out a ROBINS-I form. A provider who can name the specific reviewer and their relevant background is giving you a checkable claim. A provider who only offers "our team has extensive experience" is not.

Turnaround time: the number that's easiest to overpromise

A systematic review's timeline depends heavily on scope -- search volume, screening burden, number of outcomes, whether meta-analysis is involved -- so a fixed, generic turnaround quote before your protocol exists is itself a signal worth noticing. A provider that gives a real estimate will usually ask about your expected number of studies, your outcomes, and your deadline pressure before committing to a timeline, because a defensible estimate genuinely depends on those specifics.

Be specifically cautious of turnaround promises that look fast relative to your review's real scope. Full-text screening and data extraction are the stages most often underestimated, since screening pace depends heavily on how experienced the reviewers are with your specific inclusion criteria, and that takes real calendar time regardless of team size. A provider willing to say "we don't know yet, tell us more about your scope" before quoting a timeline is giving you a more trustworthy answer than one who quotes a number in the first email.

Confidentiality and ownership: get it in writing before you share anything

Your review, and often your unpublished data, is confidential work in progress. A provider should offer an NDA as standard practice, not as something you have to request, and the terms should be clear about who owns the final work product -- you should retain full authorship and ownership, with the provider's role limited to consulting, editing, or advisory support. Any provider whose contract language suggests co-authorship or shared ownership by default is worth reading very carefully before signing.

Pricing structure: stage-based versus hourly versus a single lump sum

Each pricing structure has a legitimate use case, but they carry different risk profiles for you as the client. Hourly billing without a cap can run over budget on a review that turns out to be larger than expected. A single lump sum for the entire project bundles the risk into one number, which is simpler but harder to verify against actual work performed at each stage. Stage-based pricing -- a defined price per phase, protocol through manuscript support -- gives you the clearest visibility into what you're paying for and lets you stop after any stage if your needs change. Ask which model a provider uses and why, and be wary of a lump sum quoted before your scope is actually defined.

Sample work and specificity of communication

Ask to see a redacted sample of prior work -- a risk-of-bias table, a PRISMA flow diagram, an extraction form -- not just a client testimonial. The actual work product tells you far more about methodological rigor than a quote does. Similarly, pay attention to how specifically a provider answers your questions during the initial conversation. Specific, checkable answers about process are a good sign. Reassurance without detail is not.

Bringing the checklist together

None of these five areas -- screening methodology, reviewer expertise, turnaround realism, confidentiality terms, and pricing structure -- are unusual or unreasonable to ask about directly. A provider confident in their own process should welcome specific questions rather than deflect them. If you find yourself getting vaguer answers the more specific your questions become, that pattern itself is useful information, independent of anything else you learn.

It's reasonable to put these same five questions to more than one provider before deciding, and to compare the specificity of the answers side by side rather than relying on memory of separate conversations. The providers worth working with are the ones whose answers get more specific, not less, the closer you look.