Food & Nutrition

Systematic Reviews for Food Safety, Novel Food Authorisation and Health Claims

In the food sector, systematic reviews are essential for safety assessment, health claims substantiation and regulatory authorisation. Puraite delivers evidence synthesis documented for EFSA applications with full traceability for European and international regulatory submissions.

AI Screening of 742 publications: 38 included, 682 excluded, 22 unclear.

Screening Details: An oat bran breakfast cereal and serum lipids in adults: a randomised crossover trial. Demo Journal of Nutrition Research · 2021 · 72 participants.

AI Decision: Unclear. AI Confidence: 62%. Assignment Status: Assigned to me.

Inclusion Criteria Met

  • Adults with normal or raised cholesterol: Adults with total cholesterol above 5.0 mmol/L.
  • Change in LDL cholesterol: The change in LDL cholesterol after four weeks is reported.

Inclusion Criteria Unclear

  • Oat beta-glucan ≥3 g per day: The daily dose of beta-glucan in the cereal is not stated.
  • Control food without beta-glucan: The control cereal is wheat-based; its fibre content is not given.

Key Evidence: “…participants ate 40 g of oat bran cereal a day or a wheat-based control cereal for four weeks…”

An AI screening run and the record it marked unclear · demo data

Supported by Europe's leading AI and life science organisations

eXISTRWTH InnovationHigh Tech NRW AcceleratorFounders FoundationGerman Acceleratorgarage33QuhrPaderborn Universitylife science factoryUniversity Hospital CologneGHA - German Health Alliance

The Challenge

Food safety and nutrition regulation demands comprehensive, reproducible evidence synthesis.

Novel Food Authorisation

Regulation (EU) 2015/2283 requires applicants to systematically compile all available safety, allergenicity and nutritional data when seeking authorisation for novel foods or novel food ingredients. EFSA expects comprehensive, reproducible literature review methodology in applications.

Health Claims Substantiation

Under Regulation (EC) No 1924/2006 a health claim must rest on generally accepted scientific evidence, judged on the totality of the available data. EFSA's NDA Panel checks how completely the application covers the pertinent studies.

Food Additive Re-Evaluation

EFSA's ongoing re-evaluation programme for food additives authorised before 2009 (under Regulation (EU) No 257/2010) requires systematic literature review to reassess safety in light of current evidence.

Pesticide Residues and Contaminants

Systematic evidence synthesis supports maximum residue level (MRL) assessments, dietary exposure modelling and risk characterisation for pesticide residues and food contaminants under Regulation (EC) No 396/2005.

How Puraite Helps

Specific workflows where Puraite accelerates regulated evidence work.

ResultStructured evidence package for EFSA novel food authorisation assessment

Comprehensive evidence synthesis for novel food applications under Regulation (EU) 2015/2283. Systematic review of safety studies, allergenicity data, toxicological evidence and nutritional information documented for EFSA applications.

Projects/ProtocolProtocolCriteria BoardSearchScreeningExtraction

Research Protocol

A systematic review of toxicity, allergenicity and nutrient data for dried yellow mealworm larvae powder, for a novel food application to EFSA.

Extract Criteria

Criteria Board

P
Population
Humans and laboratory animals
E
Exposure
Dried yellow mealworm larvae powder
C
Comparator
Diet without insect protein
O
Outcome
Toxicity, allergenicity, nutrient content

ResultHealth claim evidence with the full search and selection documented

Systematic review of human intervention studies, observational evidence and mechanistic data supporting proposed health claims. Structured per EFSA's scientific and technical guidance for health claim applications.

Projects/Risk of BiasAssessmentSummary TableTraffic Light

Tool: RoB 2 (Cochrane)6 studies

StudyD1Randomisation processD2Deviations from intended interventionsD3Missing outcome dataD4Measurement of the outcomeD5Selection of the reported resultOverall
Moreno 2018Demo Journal of Functional FoodsLLHLLH
Andersson 2019Demo Journal of Nutrition ResearchLLLLLL
Bianchi 2020Demo Journal of Lipid ResearchLLSLLS
Kowalski 2021Demo Journal of Functional FoodsSLLLLS
Fagerlund 2022Demo Journal of Clinical NutritionLLLLLL
Rasmussen 2023Demo Journal of Clinical NutritionLSLLLS
  • Legend:
  • Low risk
  • Some concerns
  • High risk
  • Not assessed

ResultEvidence-based MRL documentation with reproducible search methodology

Systematic identification and appraisal of studies on pesticide residues in food commodities. Structured data extraction for EFSA MRL assessments and dietary exposure calculations.

Projects/Search & ImportFile ImportDatabase SearchDeduplication

SourcePubMed

Drop your .ris or .nbib file hereor click to browse
TitleSourceYearStatus
Fungicide residues in wheat grain from supervised field trials...PubMed2024Imported
Processing factors for pesticide residues in milled cereals...PubMed2023Imported
Dietary exposure to fungicide residues in European diets...PubMed2022Imported

ResultAdditive safety evidence documented for EFSA applications

Systematic literature review supporting food additive safety assessments or re-evaluations. Comprehensive toxicological evidence compilation covering genotoxicity, reproductive toxicity, chronic toxicity and exposure assessment.

Projects/Data ExtractionPanel ViewTable ViewExport

Source Document

...in the 90-day study, no adverse effects were seen up to 1,000 mg/kg bw/day, the highest dose tested. Caecal enlargement was considered adaptive

3 fields linked to this passage

Extraction Form

Study design
90-day oral (OECD TG 408)
Species and strain
Wistar rat
NOAEL
1,000 mg/kg bw/day
Highest dose tested
1,000 mg/kg bw/day
Critical effect
None identified

The whole review or just the step you need.

Puraite runs systematic reviews your way: manual, hybrid or AI-assisted, with team workflows built in.

  • Role-Based Access Control

    Four project roles and the organisation's permissions decide who sees and changes what.

    Security: Role-Based Access Control
    Project Settings/Team & ReviewersTeam & ReviewersCollaborationGeneral

    Project members

    Roles control project permissions. Reviewer checkbox controls screening eligibility.

    • M. Sato

      OwnerProject Manager

      Reviewer: no

    • A. Novak

      Information Manager

      Reviewer: no

    • R. Okafor

      Collaborator

      Reviewer: yes

    • L. Brandt

      Collaborator

      Reviewer: yes

    • J. FerreiraExternal

      Viewer

      Reviewer: no

    Roles for L. Brandt

    • Project ManagerFull control of the project, members and settings.
    • Information ManagerManages publications, the extraction form and AI screening / extraction.
    • Collaborator: selectedEdits the protocol and study links; records search executions.
    • ViewerRead-only access to the project.
    Save
  • Conflict Resolution Audit

    Resolve conflicts without leaving the platform. All conflicts, reasons and resolutions are logged.

    Methodology & XAI: Conflict Resolution Audit
    Projects/Title & Abstract ScreeningDashboardMy QueueAI ScreeningResults

    Unresolved Conflicts (2)

    Oat beta-glucan in adults treated with statins: a randomised crossover trial

    Demo Journal of Lipid Research · 2021

    1 include vs 1 exclude

    Conflict
    • R. OkaforincludeRandomised trial of oat beta-glucan with LDL cholesterol as the outcome
    • L. BrandtexcludeWrong populationParticipants take lipid-lowering medication

    The population criterion excludes lipid-lowering medication. Statin users are outside the claim's target population.

    IncludeExcludeCancel

    A blend of barley and oat beta-glucan and serum lipids: a randomised controlled trial

    Demo Journal of Functional Foods · 2020

    1 include vs 1 exclude

    Conflict
    • R. OkaforexcludeWrong interventionBarley and oat beta-glucan are given together
    • L. BrandtincludeThe oat fraction alone gives 3 g per day
    Resolve Conflict
  • Risk of Bias

    Risk of bias is assessed with the tool that fits the study design, from RoB 2 and ROBINS-I to QUADAS-2 and the Newcastle-Ottawa Scale.

    Methodology & XAI: Risk of Bias
    Projects/Risk of BiasAssessmentSummary TableTraffic Light

    Tool: RoB 2 (Cochrane)6 studies

    StudyD1Randomisation processD2Deviations from intended interventionsD3Missing outcome dataD4Measurement of the outcomeD5Selection of the reported resultOverall
    Moreno 2018Demo Journal of Functional FoodsLLHLLH
    Andersson 2019Demo Journal of Nutrition ResearchLLLLLL
    Bianchi 2020Demo Journal of Lipid ResearchLLSLLS
    Kowalski 2021Demo Journal of Functional FoodsSLLLLS
    Fagerlund 2022Demo Journal of Clinical NutritionLLLLLL
    Rasmussen 2023Demo Journal of Clinical NutritionLSLLLS
    • Legend:
    • Low risk
    • Some concerns
    • High risk
    • Not assessed
  • GRADE Assessment

    Apply GRADE to evaluate evidence certainty. Generate standard Summary of Findings tables.

    Methodology & XAI: GRADE Assessment
    Projects/GRADE AssessmentEvidence ProfileSummary of Findings
    • High
    • Moderate
    • Low
    • Very low
    HTMLExport

    Summary of Findings3/7

    OutcomeStudiesParticipantsEffect Estimate (CI)CertaintyRisk with comparatorAbsolute effectPlain Language Summary
    LDL cholesterol at 4 to 8 weeks6766MD −0.24 (95% CI −0.31 to −0.17)ModerateMean 3.86 mmol/L0.24 mmol/L lower (from 0.31 lower to 0.17 lower)The intervention probably reduces LDL cholesterol at 4 to 8 weeks.
    Total cholesterol at 4 to 8 weeks5690MD −0.21 (95% CI −0.30 to −0.12)ModerateMean 5.92 mmol/L0.21 mmol/L lower (from 0.30 lower to 0.12 lower)The intervention probably reduces total cholesterol at 4 to 8 weeks.
    Gastrointestinal symptoms3416RR 1.38 (95% CI 0.94 to 2.03)Low64 per 1,00024 more per 1,000 (from 4 fewer to 66 more)The intervention may result in little to no difference in gastrointestinal symptoms.
  • PRISMA Flow Diagram

    Export the flow diagram with the record counts of every phase.

    Start a pilot: PRISMA Flow Diagram
    Projects/PRISMA Flow DiagramPRISMA 2020
    PRISMA 2020 (default)AI transparency (PRISMA-trAIce)SVGPNGExcelPowerPointWord
    IdentificationScreeningIncluded

    Records from databases:

    • PubMed (n = 388)
    • Embase (n = 471)
    • Cochrane CENTRAL (n = 169)
    • Total (n = 1,028)

    Records identified from:

    • Websites (n = 0)
    • Organisations (n = 0)
    • Citation searching (n = 0)

    Records removed before screening:

    • Duplicate records removed (n = 286)
    • Records marked as ineligible by automation tools (n = 0)
    • Records removed for other reasons (n = 0)

    Records screened

    (n = 742)

    AI-screened
    742 (100%)
    Human-screened
    742 (100%)

    Records excluded

    (n = 694)

    By AI
    682
    By human
    690
    AI agreement
    97%

    Reports sought for retrieval

    (n = 48)

    Reports not retrieved

    (n = 2)

    Reports assessed for eligibility

    (n = 46)

    AI-assessed
    46
    Human-assessed
    46

    Reports excluded

    • Wrong population (n = 8)
    • Wrong intervention (n = 11)
    • Wrong comparator (n = 6)
    • Wrong outcome (n = 7)
    • Wrong study design (n = 8)
    • Total (n = 40)
    AI-excluded
    37
    Human-excluded
    40

    Studies included in review

    (n = 6)

    • Quantitative synthesis (n = 6)
    • Qualitative synthesis (n = 0)
  • Audit trail that cannot be edited

    Decisions, overrides, comments and AI-assisted actions are attributable by user. Entries cannot be edited.

    Security: Audit trail that cannot be edited
    Title & Abstract Screening/Audit TrailRead-only

    Oat beta-glucan and LDL cholesterol in adults with mild hypercholesterolaemia: a randomised, double-blind, controlled trial

    1. AI Screening SystemIncluded, confidence 95%Decision with rationale and source quotes stored
    2. R. Okafor · Reviewer 1IncludedAdopted the AI decision after reading the abstract
    3. L. Brandt · Reviewer 2ExcludedReason given: the food matrix of the drink is not described
    4. SystemConflict openedTwo human decisions differ, record leaves the queue
    5. M. Sato · Project ManagerConflict resolved: IncludedThe dose of 3 g oat beta-glucan per day is stated in the methods
    6. SystemSent to full-text retrievalPRISMA count updated

Security and data protection

What happens to your data in Puraite

Where your research data is stored, who can reach it, what the AI providers receive and what the audit trail records.

Security
Research data stored in
EU (Paris)
Training on your content
None
Daily backups
Two EU regions
Audit trail
Read-only

AI functions send the text they work on to third-party AI providers. This can take place outside the EU, under the EU-US Data Privacy Framework or standard contractual clauses.

Preparing a novel food application or health claims dossier?

See how Puraite documents evidence synthesis for EFSA applications. Request a 30-minute walkthrough.

Funded by the German Federal Ministry for Economic Affairs and Energy and the European Union through the European Social Fund Plus (ESF Plus). Implemented by Projektträger Jülich as part of the EXIST programme.