Improving hospital readmission prediction using individualized utility analysis.

OBJECTIVE: Machine learning (ML) models for allocating readmission-mitigating interventions are typically selected according to their discriminative ability, which may not necessarily translate into utility in allocation of resources. Our objective was to determine whether ML models for allocating readmission-mitigating interventions have different usefulness based on their overall utility and discriminative ability. MATERIALS AND METHODS: We conducted a retrospective utility analysis of ML models using claims data acquired from the Optum Clinformatics Data Mart, including 513,495 commercially-insured inpatients (mean [SD] age 69 [19] years; 294,895 [57%] Female) over the period January 2016 through January 2017 from all 50 states with mean 90 day cost of $11,552. Utility analysis estimates the cost, in dollars, of allocating interventions for lowering readmission risk based on the reduction in the 90-day cost. RESULTS: Allocating readmission-mitigating interventions based on a GBDT model trained to predict readmissions achieved an estimated utility gain of $104 per patient, and an AUC of 0.76 (95% CI 0.76, 0.77); allocating interventions based on a model trained to predict cost as a proxy achieved a higher utility of $175.94 per patient, and an AUC of 0.62 (95% CI 0.61, 0.62). A hybrid model combining both intervention strategies is comparable with the best models on either metric. Estimated utility varies by intervention cost and efficacy, with each model performing the best under different intervention settings. CONCLUSION: We demonstrate that machine learning models may be ranked differently based on overall utility and discriminative ability. Machine learning models for allocation of limited health resources should consider directly optimizing for utility.

Sol-gel silica controlled release thin films for the inhibition of methicillin-resistant Staphylococcus aureus.

The incidence of methicillin-resistant Staphylococcus aureus (MRSA) infection has significantly increased. Generally, the success of this bacterium as a pathogen is attributed to its ability to adhere to surfaces and remain there, under the protection of an extracellular matrix known as biofilm. To combat MRSA with regular doses of vancomycin, efforts are continuously underway to increase its effectiveness. A promising technique is to use combinational therapeutics. In vitro experiments showed that farnesol can be used as an adjuvant with conventional antibiotics. Farnesol is a natural sesquiterpenoid and quorum-sensing molecule. The biggest obstacle to using this concept is that farnesol is highly water insoluble. This compromises its bioavailability if it were to be used along with vancomycin at the site of infection when the treatment needs to be administered in vivo. Herein we designed an efficient therapeutic strategy for the simultaneous delivery of both antibiotic and adjuvant in order to treat MRSA infections. We demonstrate that sufficient quantities of both vancomycin and farnesol can be incorporated into sol-gel silica applied as thin films on an implant surface. The incorporation of the hydrophobic farnesol does not affect the stability of the thin films and neither does it affect the controlled release of vancomycin. The data demonstrate the potent adjuvant effect of farnesol on vancomycin in inhibiting MRSA infection. In vitro experiments show the complete inhibition (10(6) fold reduction in growth compared to control) of methicillin-sensitive S. aureus (MSSA) and methicillin-resistant S. aureus (MRSA) when the ratio of vancomycin to farnesol in the sol-gel silica films is optimized. The local delivery of antibiotics minimizes the need for systemic antibiotics. The incorporation of vancomycin and farnesol into thin sol-gel films represents a new treatment paradigm for the topical delivery of antibiotics with adjuvant. The potential clinical benefits are significant and include avoiding the need for revision surgery, preventing surgical site infection and controlling healthcare costs.

Controlled release of small molecules from silica xerogel with limited nanoporosity.

Conventional sol-gel processing requires several distinct steps involving hydrolysis, condensation and drying to obtain a highly porous, glassy solid material. With the goal of achieving controlled release of small molecules, herein we focus on the acceleration of the condensation and drying steps by casting the hydrolyzed sol on a large open surface to achieve a denser 100 % silica xerogel structure. Thus, cast xerogel with a more limited porosity was prepared. The effect of synthesis parameters during sol-gel synthesis on the release kinetics of bupivacaine, vancomycin and cephalexin was investigated. The release kinetics fitted well with the Higuchi model, suggesting a diffusional release mechanism. Combining the release and nanostructure data, the formation mechanism of cast xerogel is described. Without introducing additional precursors or additives into sol-gel systems, sol-gel casting is an easy technique that further expands the applicability of sol-gel materials as excellent carriers for the controlled release of a variety of drugs.