Nicolas Griffon

Improving information retrieval using Medical Subject Headings Concepts: a test case on rare and chronic diseases.

BACKGROUND As more scientific work is published, it is important to improve access to the biomedical literature. Since 2000, when Medical Subject Headings (MeSH) Concepts were introduced, the MeSH Thesaurus has been concept based. Nevertheless, information retrieval is still performed at the MeSH Descriptor or Supplementary Concept level. OBJECTIVE The study assesses the benefit of using MeSH Concepts for indexing and information retrieval. METHODS Three sets of queries were built for thirty-two rare diseases and twenty-two chronic diseases: (1) using PubMed Automatic Term Mapping (ATM), (2) using Catalog and Index of French-language Health Internet (CISMeF) ATM, and (3) extrapolating the MEDLINE citations that should be indexed with a MeSH Concept. RESULTS Type 3 queries retrieve significantly fewer results than type 1 or type 2 queries (about 18,000 citations versus 200,000 for rare diseases; about 300,000 citations versus 2,000,000 for chronic diseases). CISMeF ATM also provides better precision than PubMed ATM for both disease categories. DISCUSSION Using MeSH Concept indexing instead of ATM is theoretically possible to improve retrieval performance with the current indexing policy. However, using MeSH Concept information retrieval and indexing rules would be a fundamentally better approach. These modifications have already been implemented in the CISMeF search engine.

A Search Engine to Access PubMed Monolingual Subsets: Proof of Concept and Evaluation in French

Background PubMed contains numerous articles in languages other than English. However, existing solutions to access these articles in the language in which they were written remain unconvincing. Objective The aim of this study was to propose a practical search engine, called Multilingual PubMed, which will permit access to a PubMed subset in 1 language and to evaluate the precision and coverage for the French version (Multilingual PubMed-French). Methods To create this tool, translations of MeSH were enriched (eg, adding synonyms and translations in French) and integrated into a terminology portal. PubMed subsets in several European languages were also added to our database using a dedicated parser. The response time for the generic semantic search engine was evaluated for simple queries. BabelMeSH, Multilingual PubMed-French, and 3 different PubMed strategies were compared by searching for literature in French. Precision and coverage were measured for 20 randomly selected queries. The results were evaluated as relevant to title and abstract, the evaluator being blind to search strategy. Results More than 650,000 PubMed citations in French were integrated into the Multilingual PubMed-French information system. The response times were all below the threshold defined for usability (2 seconds). Two search strategies (Multilingual PubMed-French and 1 PubMed strategy) showed high precision (0.93 and 0.97, respectively), but coverage was 4 times higher for Multilingual PubMed-French. Conclusions It is now possible to freely access biomedical literature using a practical search tool in French. This tool will be of particular interest for health professionals and other end users who do not read or query sufficiently in English. The information system is theoretically well suited to expand the approach to other European languages, such as German, Spanish, Norwegian, and Portuguese.

Managing free text for secondary use of health data.

OBJECTIVE To summarize the best papers in the field of Knowledge Representation and Management (KRM). METHODS A comprehensive review of medical informatics literature was performed to select some of the most interesting papers of KRM and natural language processing (NLP) published in 2013. RESULTS Four articles were selected, one focuses on Electronic Health Record (EHR) interoperability for clinical pathway personalization based on structured data. The other three focus on NLP (corpus creation, de-identification, and co-reference resolution) and highlight the increase in NLP tools performances. CONCLUSION NLP tools are close to being seriously concurrent to humans in some annotation tasks. Their use could increase drastically the amount of data usable for meaningful use of EHR.

Improving information retrieval by meta-modelling medical terminologies

This work aims at improving information retrieval in a health gateway by meta-modelling multiple terminologies related to medicine. The meta-model is based on meta-terms that gather several terms semantically related. Meta-terms, initially modelled for the MeSH thesaurus, are extended for other terminologies such as IC10 or SNOMED Int. The usefulness of this model and the relevance of information retrieval is evaluated and compared in the case of one and multiple terminologies. The results show that exploiting multiple terminologies contributes to increase recall but lowers precision.

Physician satisfaction with transition from CPOE to paper-based prescription

INTRODUCTION In January 2015, Rouen University Hospitals information system experienced serious issues. It was necessary to rapidly switch from the computerized provider order entry (CPOE) system towards a paper-based order entry (PBOE) system. This was an opportunity to evaluate prescriber opinion on the two provider order entry (POE) systems. METHODS All residents were asked to fill an augmented version of the POE satisfaction and usage survey for both POE systems. The results were compared to identify the strengths and weaknesses of each system. RESULTS Fifty-one respondents had used the CPOE system and the PBOE system. Overall, satisfaction was higher with PBOE than CPOE (odds ratio (OR)=3.74; p<0.001). Usability (OR=4.00; p<0.001), reliability (OR=8.54; p<0.001), time consumption (OR=0.50; p<0.05 - survey statement was formulated negatively), and communication with nurses (OR=14.27; p<0.0001) reached statistically better agreement. The more experience with CPOE the more residents were disillusioned with the reliability (OR=6.55; p<0.01), the usability (OR=5.68; p<0.01) and the patient safety (OR=0.27; p<0.05 - survey statement was formulated negatively) of CPOE. Although safety issues were reported for both systems, the causes were different; PBOE imposed frequent rewriting of the order while CPOE lack of usability might be unsafe. Another important issue with both POE systems was time consumption. CONCLUSION Residents did not report any increase in safety issues with the rapid switch from CPOE to PBOE. They even seemed more satisfied with the rollback to paper, which remains a possible degraded mode in case of health information technology collapse.

Influence of author's affiliation and funding sources on the results of cohort studies on occupational cancer.

BACKGROUND Reliability and credibility of research conducted by industry have been questioned, including in the field of occupational health. METHODS Cohort studies on occupational cancer published between 2000 and 2010 were compared according to their results, their conclusions, their funding, and the affiliation of their authors. RESULTS Overall, 510 articles were included. Studies published by authors with public affiliation or funded by public grants concluded that their study showed an excess of cancer more frequently (P = 0.01) than studies published by authors with private affiliation or funded by private grants (88% [95%CI = 85-91] vs. 73% [95%CI = 56-88] and 92% [95%CI = 86-97] vs. 71% [95%CI = 57-84], respectively). Discrepancies between statistical results and conclusion occurred more frequently in articles written by authors from the private sector than from the public sector (42% [IC95% = 26-60] vs. 23% [IC95% = 18-26], P = 0.02). CONCLUSIONS Industry affiliations of authors or industry support of studies are associated with the results of published studies on occupational cancer. The underlying mechanisms warrant further investigation.

Is it possible to recover personal health information from an automatically de-identified corpus of French EHRs?

De-identification aims at preserving patient confidentiality while enabling the use of clinical documents for furthering medical research. Herein, we aim to evaluate whether patient re-identification is possible on a corpus of de-identified clinical documents in French. Personal Health Identifiers are automatically marked by a de-identification system applied to the corpus, followed by reintroduction of plausible surrogates. The resulting documents are shown to individuals with varying knowledge of the documents and de-identification method. The individuals are asked to re-identify the patients. The amount of information recovered increases with familiarity with the documents and/or de-identification method. Surrogate re-introduction with localization from the same (vs. different) geographical area as the original documents is found more effective. The amount of information recovered was not sufficient to re-identify any of the patients, except when privileged access to the hospital health information system and several documents about the same patient were available.

Evaluating alignment quality between iconic language and reference terminologies using similarity metrics.

BackgroundVisualization of Concepts in Medicine (VCM) is a compositional iconic language that aims to ease information retrieval in Electronic Health Records (EHR), clinical guidelines or other medical documents. Using VCM language in medical applications requires alignment with medical reference terminologies. Alignment from Medical Subject Headings (MeSH) thesaurus and International Classification of Diseases – tenth revision (ICD10) to VCM are presented here. This study aim was to evaluate alignment quality between VCM and other terminologies using different measures of inter-alignment agreement before integration in EHR.MethodsFor medical literature retrieval purposes and EHR browsing, the MeSH thesaurus and the ICD10, both organized hierarchically, were aligned to VCM language. Some MeSH to VCM alignments were performed automatically but others were performed manually and validated. ICD10 to VCM alignment was entirely manually performed. Inter-alignment agreement was assessed on ICD10 codes and MeSH descriptors, sharing the same Concept Unique Identifiers in the Unified Medical Language System (UMLS). Three metrics were used to compare two VCM icons: binary comparison, crude Dice Similarity Coefficient (DSCcrude), and semantic Dice Similarity Coefficient (DSCsemantic), based on Lin similarity. An analysis of discrepancies was performed.ResultsMeSH to VCM alignment resulted in 10,783 relations: 1,830 of which were manually performed and 8,953 were automatically inherited. ICD10 to VCM alignment led to 19,852 relations. UMLS gathered 1,887 alignments between ICD10 and MeSH. Only 1,606 of them were used for this study. Inter-alignment agreement using only validated MeSH to VCM alignment was 74.2% [70.5-78.0]CI95%, DSCcrude was 0.93 [0.91-0.94]CI95%, and DSCsemantic was 0.96 [0.95-0.96]CI95%. Discrepancy analysis revealed that even if two thirds of errors came from the reviewers, UMLS was nevertheless responsible for one third.ConclusionsThis study has shown strong overall inter-alignment agreement between MeSH to VCM and ICD10 to VCM manual alignments. VCM icons have now been integrated into a guideline search engine (http://www.cismef.org) and a health terminologies portal (http://www.hetop.eu).

Surgical hand rubbing: knowledge and declared use of health care workers.

of the 5 swabs tested. No samples were positive for MRSA from any patient/slit lamp combination. The range of bacteria cultured from slit lamps included common commensal bacteria such as CNS but also environmental organisms, including Bacillus species and Rhodococcus species, the latter organism being of growing clinical significance. Once contaminated, such gram-positive organisms may survive for extended periods of time because of their relative resistance to dessication. The results of this study demonstrated no bacterial growth occurred when the slit lamp was vigorously wiped twice with alcohol, indicating that this is an effective method of bacterial sterilization of slit lamps in clinical practice. We therefore recommend that alcohol cleaning regimes should be introduced as normal clinical practice within the ophthalmic hospital arena, as an effective tool for minimising potential bacterial cross contamination between patients.

Impact on medical practice of accessing pharmaceutical records

INTRODUCTION The pharmaceutical record system (PRS) is a French nationwide centralized electronic database shared among all community pharmacists listing all drugs dispensed by community pharmacists in the last four months. The objective of this study, the Medication Assessment Through Real time Information eXchange - Distributed Pharmaceutical Record System (MATRIX - DPRS) study, was to assess the clinical impact of the PRS upon granting access to physicians in three hospital specialties: anesthesiology, emergency medicine and geriatrics. MATERIAL AND METHODS A multicenter prospective study was conducted in six hospital departments, two per specialty. Participating physicians noted medication information found exclusively in the pharmaceutical record (PR) of each patient unavailable elsewhere and any diagnostic or therapeutic management changes resulting from the PR information. The primary objective was to assess the proportion of diagnostic or therapeutic management changes attributable to the PR among patients who had an accessible PR. RESULTS The inclusion level ranged from 1.1 to 30% in the six departments. The rate of diagnostic or therapeutic management changes was highest in geriatrics (n = 31/67; 46.3% 95% Confidence IntervaI (CI): 34.0-58.9%) and lowest in anesthesiology (n = 36/227; 15.9% 95% CI: 11.4-21.3%). Emergency medicine was intermediate (n = 5/22; 22.7% 95% CI: 7.8-45.4%). CONCLUSION Although the inclusion rate and statistical precision were low, these findings suggest that the information contained in the PRS is useful and may result in modifying patient management in a sizeable proportion of patients. This opens the prospect of evaluating other hospital specialties, as well as primary and secondary care settings.