Applying FAIR Principles to Plant Phenotypic Data Management in GnpIS

GnpIS is a data repository for plant phenomics that stores whole field and greenhouse experimental data including environment measures. It allows long-term access to datasets following the FAIR principles: Findable, Accessible, Interoperable, and Reusable, by using a flexible and original approach. It is based on a generic and ontology driven data model and an innovative software architecture that uncouples data integration, storage, and querying. It takes advantage of international standards including the Crop Ontology, MIAPPE, and the Breeding API. GnpIS allows handling data for a wide range of species and experiment types, including multiannual perennial plants experimental network or annual plant trials with either raw data, i.e., direct measures, or computed traits. It also ensures the integration and the interoperability among phenotyping datasets and with genotyping data. This is achieved through a careful curation and annotation of the key resources conducted in close collaboration with the communities providing data. Our repository follows the Open Science data publication principles by ensuring citability of each dataset. Finally, GnpIS compliance with international standards enables its interoperability with other data repositories hence allowing data links between phenotype and other data types. GnpIS can therefore contribute to emerging international federations of information systems.

Why is antimicrobial de-escalation under-prescribed for urinary tract infections?

PURPOSE

To assess the frequency of de-escalation in inpatients treated for community-acquired urinary tract infection and the frequency of conditions legitimating not de-escalating therapy.

METHODS

A retrospective study of inpatients (age >15 years) at a large academic hospital who were empirically treated for urinary tract infections due to Escherichia coli susceptible to at least one of the following antibacterial agents: amoxicillin, co-amoxiclav, and cotrimoxazole. De-escalation was defined as the replacement of the empirical broad-spectrum therapy by amoxicillin, co-amoxiclav, or cotrimoxazole.

RESULTS

Eighty patients were included. De-escalation was prescribed for 32 of 69 patients for whom it was possible from both a bacteriological and clinical point of view (46 %, 95 % CI, 34-59 %). Initial treatment was switched to amoxicillin (n = 21), co-amoxiclav (n = 2), or cotrimoxazole (n = 8). Thirteen conditions justifying not de-escalating antibacterial therapy were detected in 11 of 48 patients who were not de-escalated (23 %, 95 % CI, 12-37 %): shock, n = 5; renal abscess, n = 1; obstructive uropathy, n = 4; bacterial resistance or clinical contraindication to both cotrimoxazole and β-lactams, n = 3.

CONCLUSIONS

De-escalation is under-prescribed for urinary tract infections. Omission of de-escalation is seldom legitimate. Interventions aiming to de-escalate antibacterial therapy for UTIs should be actively implemented.