Deep learning classification of drug-related problems from pharmaceutical interventions issued by hospital clinical pharmacists during medication prescription review: a large-scale descriptive retrospective study in a French university hospital.

OBJECTIVES: Pharmaceutical interventions are proposals made by hospital clinical pharmacists to address sub-optimal uses of medications during prescription review. Pharmaceutical interventions include the identification of drug-related problems, their prevention and resolution. The objective of this study was to exploit a newly developed deep neural network classifier to identify drug-related problems from pharmaceutical interventions and perform a large retrospective descriptive analysis of them in a French university hospital over a 3-year period. METHODS: Data were collected from prescription support software from 2018 to 2020. A classifier running in Python 3.8 and using Keras library was then used to automatically categorise drug-related problems from pharmaceutical interventions according to the coding of the French Society of Clinical Pharmacy. RESULTS: 2 930 656 prescription lines were analysed for a total of 119 689 patients. Among these prescription lines, 153 335 (5.2%) resulted in pharmaceutical interventions (n=48 202 patients; 40.2%). Pharmaceutical interventions were predominantly observed in patients aged 65 years or older (n=26 141 patients out of 53 186; 49.1%) and in patients taking five or more medications (44 702 patients out of 93 419; 47.8%). The most frequently identified types of drug-related problems associated with pharmaceutical interventions were 'Non-conformity to guidelines or contra-indication' (n=88 523; 57.7%), 'Overdosage' (16 975; 11.1%) and 'Improper administration' (13 898; 9.1%). The most frequently encountered drugs were: paracetamol (n=10 585; 6.9%), esomeprazole (6031; 3.9%), hydrochlorothiazide (2951; 1.9%), enoxaparin (2191; 1.4%), tramadol (1879; 1.2%), calcium (2073; 1.3%), perindopril (1950; 1.2%), amlodipine (1716; 1.1%), simvastatin (1560; 1.0%) and insulin (1019; 0.7%). CONCLUSIONS: The deep neural network classifier used met the challenge of automatically classifying drug-related problems from pharmaceutical interventions from a large database without mobilising significant human resources. The use of such a classifier can lead to alerting caregivers about certain risky practices in prescription and administration, and triggering actions to improve patients' therapeutic outcomes.

Complication Rates Are Not Higher After Outpatient Versus Inpatient Fast-Track Total Knee Arthroplasty: A Propensity-Matched Prospective Comparative Study.

BACKGROUND: This prospective study aimed to compare the complication rates and clinical outcomes of propensity-matched patients who received fast-track total knee arthroplasty (FT TKA) in outpatient versus inpatient settings. METHODS: Patients (n = 629) who received FT TKA at various outpatient (n = 176) and inpatient (n = 462) surgery rates were prospectively followed until 90 days after surgery. The decision between inpatient versus outpatient FT TKA was made on a case-by-case basis, depending on consultation between the surgeon and patient. Complications were collected to distinguish between intraoperative complications, complications with no readmission, complications with readmission, and complications with reoperation. Propensity scores based on age, sex, body mass index, and the American Society of Anesthesiologists score were used to match outpatient to inpatient FT TKA. A cumulative incidence function was computed by taking the time to diagnose any postoperative complication in the first 90 days as the end point. RESULTS: Propensity score matching (1:2 ratio) for comparison resulted in 173 outpatient FT TKAs and 316 inpatient FT TKAs. No significant differences were observed between outpatient versus inpatient FT TKA for intraoperative complication rates (2% in both groups). At 90-day follow-up, no significant differences were observed between outpatient versus inpatient FT TKA for total complications with no readmission (8.0 versus 7.9%), complications with readmission but no reoperation (1.1 versus 0.6%), and complications with reoperation (4.0 versus 4.4%). A comparison of postoperative complication diagnosis time using the cumulative incidence function revealed no significant differences between outpatient versus inpatient FT TKA. CONCLUSIONS: The present study revealed that there were no differences in 90-day postoperative complication rates between outpatient and inpatient FT TKA and that there were also no differences in rates of intraoperative complications, readmissions, or reoperations. These findings may encourage hesitant surgeons to move toward outpatient TKA pathways, as there is no greater risk of early postoperative complications that could be more difficult to manage after discharge. LEVEL OF EVIDENCE: Level II.

PvdL Orchestrates the Assembly of the Nonribosomal Peptide Synthetases Involved in Pyoverdine Biosynthesis in Pseudomonas aeruginosa.

The pyoverdine siderophore is produced by Pseudomonas aeruginosa to access iron. Its synthesis involves the complex coordination of four nonribosomal peptide synthetases (NRPSs), which are responsible for assembling the pyoverdine peptide backbone. The precise cellular organization of these NRPSs and their mechanisms of interaction remain unclear. Here, we used a combination of several single-molecule microscopy techniques to elucidate the spatial arrangement of NRPSs within pyoverdine-producing cells. Our findings reveal that PvdL differs from the three other NRPSs in terms of localization and mobility patterns. PvdL is predominantly located in the inner membrane, while the others also explore the cytoplasmic compartment. Leveraging the power of multicolor single-molecule localization, we further reveal co-localization between PvdL and the other NRPSs, suggesting a pivotal role for PvdL in orchestrating the intricate biosynthetic pathway. Our observations strongly indicates that PvdL serves as a central orchestrator in the assembly of NRPSs involved in pyoverdine biosynthesis, assuming a critical regulatory function.

Outcomes of total elbow arthroplasty in the treatment of distal humeral fractures in the elderly: a retrospective cohort comparison between primary arthroplasty and arthroplasty secondary to failed internal fixation.

BACKGROUND: The primary objective of this study was to compare the clinical outcomes of total elbow arthroplasty as the index procedure in the treatment of traumatic distal humerus fractures with those of secondary total elbow arthroplasty after failed internal fixation. The secondary objective was to compare the complication rates and the radiographic results in the 2 groups. Our hypothesis was that the clinical results of total elbow arthroplasty performed after failed internal fixation were comparable to those of primary total elbow arthroplasty in the treatment of distal humerus fractures in the elderly population. METHODS: We conducted a retrospective cohort comparison study, including 60 patients with a median age of 80 years (71-85 years), who either underwent a primary total elbow arthroplasty (group 1; 45 patients) or secondary total elbow arthroplasty after failed internal fixation (group 2; 15 patients) in the treatment of a post-traumatic supra and intercondylar fracture of the distal humerus, between January 2004 and January 2021. The clinical examination, including the Mayo Elbow Performance Score and triceps proficiency test, complication rates, and the need for reoperation were noted. The average clinical and radiographic follow-up was 40.8 months (24-120 months). RESULTS: The clinical results of the 2 groups were comparable when looking at the Mayo Elbow Performance Score (90.00 [85.00, 100.00], P = .486). With regard to complications, there were 2 surgical site infections in group 1 and 3 in group 2 (P = .099), 1 case of mechanical loosening of the humeral component in group 1 and 1 in group 2 (P = .448), and 1 patient with triceps insufficiency in group 1. CONCLUSIONS: Secondary total elbow arthroplasty after failed internal fixation has shown good functional results and a complication rate comparable to that of index total elbow arthroplasty in the treatment of articular fractures of the distal humerus in the elderly.

Spectral Phasor Applied to Spectrally-Resolved Single Molecule Localization Microscopy.

Spectrally-resolved single-molecule localization microscopy (srSMLM) has emerged as a powerful tool for exploring the spectral properties of single emitters in localization microscopy. By simultaneously capturing the spatial positions and spectroscopic signatures of individual fluorescent molecules, srSMLM opens up the possibility of investigating an additional dimension in super-resolution imaging. However, appropriate and dedicated tools are required to fully capitalize on the spectral dimension. Here, we propose the application of the spectral phasor analysis as an effective method for summarizing and analyzing the spectral information obtained from srSMLM experiments. The spectral phasor condenses the complete spectrum of a single emitter into a two-dimensional space, preserving key spectral characteristics for single-molecule spectral exploration. We demonstrate the effectiveness of spectral phasor in efficiently classifying single Nile Red fluorescence emissions from largely overlapping cyanine fluorescence signals in dual-color PAINT experiments. Additionally, we employed spectral phasor with srSMLM to reveal subtle alterations occurring in the membrane of Gram-positive Enterococcus hirae in response to gramicidin exposure, a membrane-perturbing antibiotic treatment. Spectral phasor provides a robust, model-free analytic tool for the detailed analysis of the spectral component of srSMLM, enhancing the capabilities of multi-color spectrally-resolved single-molecule imaging.

Rhinoconjunctivitis severity induced by cat exposure influences early and late asthmatic responses: Evidence from an environmental exposure chamber.

BACKGROUND: The impact of allergic rhinoconjunctivitis on the early (EAR) and late asthmatic response (LAR) has yet to be assessed during optimal allergen exposure conditions. OBJECTIVE: We aimed to assess predictive factors of the EAR and LAR and to evaluate the relation between rhinitis, conjunctivitis and asthma induced by cat allergen exposure in an environmental exposure chamber (EEC). METHODS: Data from two cohort studies involving asthmatic patients with cat allergy who performed a cat allergen exposure challenge in ALYATEC EEC were analysed. Spirometry, visual analogue scale (VAS) for asthma, VAS for rhinitis, Total Nasal Symptoms Score, Total Ocular Symptoms Score (TOSS), Rhinoconjunctivitis Total Symptoms Score and Abelson score were used to assess asthma, rhinitis and conjunctivitis during and after exposure. RESULTS: An EAR occurred in 65.1% of patients, 32.1% of whom had a LAR. The diameter of the prick test to cat allergens and non-specific bronchial hypersensitivity level were independent risk factors for EAR (p < .05). No independent risk factors for LAR were identified. Rhinoconjunctivitis severity during exposure correlated with the asthma VAS during EAR and LAR (p < .05). Allergen exposure time needed to trigger an EAR correlated with the Abelson score during exposure (p < .05). The asthma VAS and TOSS during exposure correlated with faster LAR occurrence (p < .05). CONCLUSION: Prick test size and non-specific bronchial hypersensitivity level were confirmed as independent predictive factors of EAR during allergen exposure in an EEC. This study demonstrated the relation between the severity of rhinitis, conjunctivitis and asthma induced by allergen exposure for both EAR and LAR.

Targeted Single Particle Tracking with Upconverting Nanoparticles.

Single particle tracking (SPT) is a powerful technique for real-time microscopic visualization of the movement of individual biomolecules within or on the surface of living cells. However, SPT often suffers from the suboptimal performance of the photon-emitting labels used to tag the biomolecules of interest. For example, fluorescent dyes have poor photostability, while quantum dots suffer from blinking that hampers track acquisition and interpretation. Upconverting nanoparticles (UCNPs) have recently emerged as a promising anti-Stokes luminescent label for SPT. In this work, we demonstrated targeted SPT using UCNPs. For this, we synthesized 30 nm diameter doped UCNPs and coated them with amphiphilic polymers decorated with polyethylene glycol chains to make them water-dispersible and minimize their nonspecific interactions with cells. Coated UCNPs highly homogeneous in brightness (as confirmed by a single particle investigation) were functionalized by immunoglobulin E (IgE) using a biotin-streptavidin strategy. Using these IgE-UCNP SPT labels, we tracked high-affinity IgE receptors (FcεRI) on the membrane of living RBL-2H3 mast cells at 37 °C in the presence and absence of antigen and obtained good agreement with the literature. Moreover, we used the FcεRI-IgE receptor-antibody system to directly compare the performance of UCNP-based SPT labels to organic dyes (AlexaFluor647) and quantum dots (QD655). Due to their photostability as well as their backgroundless and continuous luminescence, SPT trajectories obtained with UCNP labels are no longer limited by the photophysics of the label but only by the dynamics of the system and, in particular, the movement of the label out of the field of view and/or focal plane.

Deep learning application to automated classification of recommendations made by hospital pharmacists during medication prescription review.

PURPOSE: Recommendations to improve therapeutics are proposals made by pharmacists during the prescription review process to address suboptimal use of medicines. Recommendations are generated daily as text documents but are rarely reused beyond their primary use to alert prescribers and caregivers. If recommendation data were easier to summarize, they could be used retrospectively to improve safeguards for better prescribing. The objective of this work was to train a deep learning algorithm for automated recommendation classification to valorize the large amount of recommendation data. METHODS: The study was conducted in a French university hospital, at which recommendation data were collected throughout 2017. Data from the first 6 months of 2017 were labeled by 2 pharmacists who assigned recommendations to 1 of the 29 possible classes of the French Society of Clinical Pharmacy classification. A deep neural network classifier was trained to predict the class of recommendations. RESULTS: In total, 27,699 labeled recommendations from the first half of 2017 were used to train and evaluate a classifier. The prediction accuracy calculated on a validation dataset was 78.0%. We also predicted classes for unlabeled recommendations collected during the second half of 2017. Of the 4,460 predictions reviewed, 67 required correction. When these additional labeled data were concatenated with the original dataset and the neural network was retrained, accuracy reached 81.0%. CONCLUSION: To facilitate analysis of recommendations, we have implemented an automated classification system using deep learning that achieves respectable performance. This tool can help to retrospectively highlight the clinical significance of daily medication reviews performed by hospital clinical pharmacists.