Time-saving study :
Comparison on scientific literature review processes between manual method and the NOÉ application
Juillet 29th 2025
Study objective
Methodology
Detailed Results
Results Analysis
Conclusion
How does NOÉ help save time ?
Study objective
This study aims to objectively measure the potential time savings provided by NOÉ compared to a traditional manual method for scientific literature
review in the context of clinical evaluation of medical devices.
Significant learning effect: NOÉ was used for the first time by evaluators during the hemodialysis review, achieving a 12.6% time saving. During
the second review (diabetes), users who had already used the tool showed much higher time savings (54.2%). These results highlight a strong
learning curve, with significant gains from the first use and a cumulative effect during the second review.
Methodology
2
Literature reviews
114
Articles evaluated in total
6
Participating evaluators
2
Methods compared
NOÉ Method
Each evaluator processes articles using the NOÉ web application, designed to automate the sorting and labeling process of articles
through artificial intelligence.
through artificial intelligence.
Manual Method
Articles are processed manually using a pre-prepared Excel spreadsheet, reproducing the traditional evaluation process.
Evaluator profiles : 5 beginner evaluators (<1 year of experience) and 1 expert evaluator (>3 year of experiencce) participated in the study.
Evaluators 6 and 7 were excluded from the analysis for non-compliance with instructions.
Detailed Results
Processing time by evaluator
| Evaluator | Review #1 NOÉ (min) | Review #1 Manual (min) | Review #2 NOÉ (min) | Review #2 Manual (min) |
|---|---|---|---|---|
| Evaluator 1 | 12 | 30 | 7.5 | 15 |
| Evaluator 2 | — | — | 9 | 20 |
| Evaluator 4 | 15 | 32 | 10 | 21 |
| Evaluator 10 | 6 | 14 | 6 | 15 |
* Evaluator 2 is not included for review #1 (hemodialysis) as instructions were not followed.
Comparative analysis by review
Average time by method and review (minutes)
Review n°1 : Hemodialysis
- Average NOÉ time : 11,0 min
- Average Manual time : 25,33 min
- Time savings : 56,6%
Review n°2 : Diabetes
- Average NOÉ time : 8,13 min
- Average Manual time : 17,75 min
- Time savings : 54,2%
55.4%
Overall average time savings
9.6min
Average NOÉ time
21.5min
Average Manual time
Results Analysis
The results show significant variability between evaluators and between reviews, which can be explained by :
- The experience level of evaluators (predominantly beginners)
- The learning curve associated with using NOÉ
- Variable complexity of articles according to therapeutic domains
- Progressive adaptation of users to NOÉ’s features
Rapid learning effect from first use : Although evaluators discovered NOÉ during the first review (hemodialysis), they already achieved a
notable time saving of 56.6%. During the second review (diabetes), this gain was maintained at 54.2%, demonstrating rapid learning and
consistent performance. These results illustrate quick user skill development and confirm that NOÉ integrates effectively into practices from the
first review.
Exclusion of one evaluator in review #1 : For hemodialysis, evaluator 2 was not included as instructions were not followed. Without them, the
time saving reaches 56.6%, showing that NOÉ already provides significant benefits from the first correctly completed review.
Conclusion
NOÉ demonstrates significant potential for improving the efficiency of scientific literature reviews, with an average time saving of 55.4% across all reviews.
From the first review, users saved time thanks to NOÉ. The second review confirmed the stability of these performances, with consistent
gains after initial exposure to the tool. This suggests that NOÉ enables immediate and lasting gains.
From the first review, users saved time thanks to NOÉ. The second review confirmed the stability of these performances, with consistent
gains after initial exposure to the tool. This suggests that NOÉ enables immediate and lasting gains.
Time savings on 2nd review – Rapid mastery of NOÉ
How does NOÉ help save time?
Intelligent automation
NOÉ’s machine learning algorithms automatically analyze articles and propose pre-sorting based on defined eligibility criteria.
Natural language processing
NOÉ’s machine learning algorithms automatically analyze articles and propose pre-sorting based on defined eligibility criteria.
Optimized interface
NOÉ’s web interface is designed to minimize clicks and optimize the review workflow, reducing evaluators’ cognitive load.
