Scientific review · Evidence synthesis · Project evaluation

A clear assessment of what the evidence supports.

Independent scientific review for manuscripts, technical decisions and research projects. We examine the claims, methods and supporting evidence, distinguish what is demonstrated from what remains uncertain, and identify what needs to change—or what can confidently stand.

Discuss a review ↗ See example assessments ↓

From a focused second opinion to a structured evaluation.

Experience behind the assessment

XpertScientific brings together scientists with experience in research, methodological assessment, manuscript review and research evaluation across multiple fields. Members of our team have served as journal reviewers, expert evaluators and rapporteurs for EU MSCA, Horizon Europe Cluster 5 and Mission proposals, and as external evaluators and monitors of Horizon Europe research projects. We match each assignment to the scientific expertise it requires. Meet our scientists ↗

Some of our clients

Client logos

When an independent view helps

Three ways to bring clarity to the evidence.

Bring us a manuscript with a disputed conclusion, a body of literature that needs interpreting, or project results that require independent assessment. We agree the question and scope before the review begins.

01

Scientific review & second opinions

Resolve substantive concerns about the methods, interpretation or conclusions in a manuscript or technical report.

  • Are the methods appropriate to the question?
  • Do the results support the conclusions?
  • Which concerns require revision, further analysis or additional evidence?
You provide
Manuscript or report, supporting material and any specific concerns.
You receive
An annotated document and review report with prioritised findings and a clear distinction between defensible and unsupported claims.
02

Evidence review & synthesis

Establish what published research can tell you about a defined question—and how well it applies to your situation.

  • Which sources are relevant, robust and applicable?
  • Why do findings differ between studies or settings?
  • What does the evidence justify in the context of your decision?
You provide
The decision or research question, its context and any existing sources or evidence base.
You receive
A documented synthesis, evidence table or matrix, and a clear account of what the evidence supports, where uncertainty remains and how well the findings apply to your situation.
03

Research project evaluation

Assess scientific progress and reported achievements against the project’s agreed objectives, criteria and intended use.

  • What has been delivered and what has actually been demonstrated?
  • Which claims are supported, qualified or still unverified?
  • What follow-up is justified by the original objectives?
You provide
Objectives, evaluation criteria, deliverables, reports and supporting results.
You receive
An evaluation report showing the evidence behind each judgement, where gaps remain and what follow-up would be justified.

If you need us to check underlying data, reproduce analyses or carry out additional research, we can add this to the assignment.

Scientific judgement in practice

From evidence to a defensible conclusion.

consulting example image

Can a wetland restoration plan meet its catchment nitrogen target?

Hypothetical example

The question

A restoration proposal targets a 30% reduction in annual total-nitrogen export. It cites studies reporting around 50% removal of nitrogen entering a wetland. The proposed wetland is expected to intercept about 35% of the catchment’s annual nitrogen load; the remainder follows other flow paths.

Wetland restoration field assessment

First check: can the claimed effect be achieved at this scale?

Catchment nitrogen load intercepted Assumed removal of intercepted load Simplified catchment reduction

This is an illustrative mass-balance check, not a site prediction. It assumes the same nitrogen measure and annual period, transferable removal efficiency, and no compensating changes elsewhere in the catchment.

Evidence examined

Reported total-nitrogen loads and removal efficiencies, hydraulic loading and retention conditions, wetland area and design, seasonal and high-flow coverage, water balances, bypass flows and the fraction of the local load actually routed through the proposed wetland.

Assessment

The cited efficiency does not support the 30% catchment target as currently framed. Under the proposal’s own interception and removal assumptions, the simplified reduction is about 17.5%. The literature efficiency is also not a fixed transfer factor: wetland performance varies with loading and hydrology, and incomplete water balances or missed high-flow events can bias apparent removal. The intercepted fraction and local operating conditions therefore need to be checked for the proposed site.

Practical consequence

Revise the expected benefit or assess designs and locations that intercept more of the load, potentially alongside complementary measures. Before committing to a performance target, prioritise a site-specific water balance and flow-weighted nitrogen-load measurements that capture the events responsible for a substantial share of annual export.

Review output

An evidence matrix separating transferable findings from local assumptions, a checked catchment-scale load calculation and a plan for the local measurements and checks still needed.

Scientific context: Land et al. (2016), systematic review of nutrient removal in created and restored freshwater wetlands. The percentages in this hypothetical scenario are invented for illustration and are not taken from the cited review.

Does the manuscript support a claim of slowed disease progression?

Hypothetical example

The question

A manuscript combines laboratory findings with a short, uncontrolled patient study. A biomarker associated with prognosis changes after treatment and symptoms improve. The authors conclude that the treatment slows disease progression, although progression was not directly measured and the biomarker is not validated as a surrogate for treatment benefit in this setting.

Medical consulting example image

Separate the levels of the claim

Biological mechanismLaboratory results may support plausibility; experimental controls and alternative explanations still matter.
Observed patient changesBiomarker and symptom changes are reported, but treatment causation remains uncertain without a suitable comparator.
Disease modificationNeither directly measured nor established through a validated surrogate endpoint in this treatment context.

Evidence examined

Experimental controls, patient selection, comparator availability, outcome definitions and timing, follow-up, missing data, and the evidence defining the biomarker’s context of use—including whether it is simply prognostic or whether treatment-induced changes in the biomarker have been validated as predictors of clinical benefit.

Assessment

The disease-modification conclusion is unsupported by the described study. The observed changes may still be scientifically useful and can be reported with appropriate limitations. However, an uncontrolled design does not establish that treatment caused them, and a biomarker associated with prognosis does not by itself establish that changing the biomarker predicts clinical benefit or slowed progression.

Practical consequence

Revise the title, abstract and discussion so that mechanistic plausibility, observed symptom or biomarker changes and disease modification are treated as separate claims. Identify the comparative evidence, progression outcomes and follow-up needed to test the stronger conclusion.

Review output

An annotated manuscript and claim-by-claim assessment distinguishing defensible statements, revisions required and questions needing further evidence.

Scientific context: US FDA resources on biomarkers, clinical outcomes and surrogate endpoints. This example illustrates manuscript assessment, not treatment advice.

What has a pilot warning system demonstrated—and what remains untested?

Hypothetical example

The question

A research project integrates sensors, a predictive model and a water-quality warning interface. The agreed integration milestone has been demonstrated at two pilot sites. Warning performance meets the project target in routine conditions, but high-flow events are sparse, some evaluation data informed model tuning, and performance at sites excluded from development has not been established.

Assess each criterion separately

Demonstrated System integration Records support end-to-end operation at the two pilot sites.
Supported with limitations Pilot warning performance The target is met in routine tested conditions, but high-flow coverage is sparse and development/evaluation separation is imperfect.
Unverified Transfer to new sites No validation has been completed at sites excluded from model development and tuning.

Evidence examined

Agreed objectives and acceptance criteria, integration records, sensor quality controls, model-development and evaluation datasets, event coverage, warning detection and false-alarm performance, behaviour during data gaps or sensor drift, and operational responsibilities for maintenance and data access.

Assessment

The integration milestone is met, and the available evidence supports warning performance in the tested routine conditions; the broader readiness claim remains unverified. Sparse high-flow observations and incomplete separation between development and evaluation weaken the evidence for generalisation. Absence of new-site validation is an evidence gap, not a demonstrated system failure, and any additional deployment requirements must be distinguished from the project’s original commitments.

Practical consequence

Retain the positive milestone finding. Plan a focused validation exercise using data and sites excluded from development, with relevant high-flow events and pre-specified warning-performance criteria. Clarify operational responsibilities before extending the claim to wider routine use.

Review output

An evaluation matrix showing which achievements are demonstrated, which findings need qualification, where evidence is missing and what follow-up is appropriate.

Scientific context: European Commission technology-readiness definitions distinguish demonstration in relevant environments from proof in operational use. No formal technology-readiness level is assigned in this illustrative assessment.

From findings to action

A review you can use and scrutinise.

You receive a clear written assessment showing how the conclusions follow from the evidence, together with any changes or follow-up we recommend. We agree the format and level of detail before work begins.

Specific, prioritised findings

What is supported, what needs revision and which gaps materially affect the conclusion.

Evidence behind each judgement

We point to the relevant sources, results or criteria and make clear any assumptions or limits on what we were able to check.

Practical next steps

Clear revisions or follow-up work, with essential actions separated from useful further development.

Findings explained in context

A discussion of the report can be included so your team understands the implications and remaining questions.

How it works

From the question to a clear assessment.

01 / DEFINE

Clarify the question

Tell us what needs assessing, who will use the findings and your deadline.

02 / AGREE

Set the scope

We agree what we will review, who needs to be involved, what you will receive, and the fee and timetable.

03 / EXAMINE

Review the evidence

We review the material, check the evidence behind the key claims and come back to you if anything needs clarification.

04 / REPORT

Explain the findings

You receive the agreed report or assessment, with the main findings, limitations and recommended next steps.

Before we begin

Practical questions.

A useful review starts with a clear question. Before work begins, we confirm that we have the right expertise and address confidentiality and any potential conflicts.

How is scientific review different from language editing?

Scientific review examines the methods, evidence and interpretation: whether the conclusions are justified and what needs substantive revision. Language editing improves expression and readability. We can provide both services together, with a clear separation between the scientific review and the language editing.

Can you review work without the underlying data?

Yes. A document-based review can assess the scientific argument, reported methods and consistency of the conclusions. It cannot establish that unexamined data or calculations are correct. The report states what was available, what was checked and which questions require data or code access.

Do you conduct systematic reviews or meta-analyses?

The review method depends on the question, the evidence available and the resources required. We do not describe a targeted evidence review as a systematic review. If you need a systematic review or meta-analysis, we first assess feasibility and agree the protocol, search and selection methods, appraisal, synthesis approach and specialist input needed.

How independent is the assessment?

We identify prior involvement and potential conflicts before accepting the assignment. The assessment is based on the agreed question and available evidence; its conclusions are not predetermined. Where we have helped develop the work, that involvement is disclosed and the review is not presented as independent of our contribution.

Can you evaluate project impact?

We can assess the evidence behind claimed outcomes and impacts where this falls within our scientific expertise. Delivering an output, observing a change and demonstrating that a project caused that change require different evidence. We will tell you which questions the available information can answer and where specialist impact-evaluation methods may be needed.

How are fees, timing and confidentiality handled?

The fee and timetable depend on the question, volume and complexity of the material, and depth of checking required. We agree these before work begins, together with access and confidentiality arrangements. Any proposed extension of scope is discussed with you first.

Start with the question

What needs an independent assessment?

Send a brief outline of the work and the decision or conclusion it needs to inform. We will confirm whether we have the right expertise and suggest a suitable scope.

Discuss a review ↗