A Neurosurgical Readmissions Reduction Program in an Academic Hospital Leveraging Machine Learning, Workflow Analysis, and Simulation.

BACKGROUND: Predicting 30-day hospital readmissions is crucial for improving patient outcomes, optimizing resource allocation, and achieving financial savings. Existing studies reporting the development of machine learning (ML) models predictive of neurosurgical readmissions do not report factors related to clinical implementation. OBJECTIVES: Train individual predictive models with good performance (area under the receiver operating characteristic curve or AUROC > 0.8), identify potential interventions through semi-structured interviews, and demonstrate estimated clinical and financial impact of these models. METHODS: Electronic health records were utilized with five ML methodologies: gradient boosting, decision tree, random forest, ridge logistic regression, and linear support vector machine. Variables of interest were determined by domain experts and literature. The dataset was split divided 80% for training and validation and 20% for testing randomly. Clinical workflow analysis was conducted using semi-structured interviews to identify possible intervention points. Calibrated agent-based models (ABMs), based on a previous study with interventions, were applied to simulate reductions of the 30-day readmission rate and financial costs. RESULTS: The dataset covered 12,334 neurosurgical intensive care unit (NSICU) admissions (11,029 patients); 1,903 spine surgery admissions (1,641 patients), and 2,208 traumatic brain injury (TBI) admissions (2,185 patients), with readmission rate of 13.13, 13.93, and 23.73%, respectively. The random forest model for NSICU achieved best performance with an AUROC score of 0.89, capturing potential patients effectively. Six interventions were identified through 12 semi-structured interviews targeting preoperative, inpatient stay, discharge phases, and follow-up phases. Calibrated ABMs simulated median readmission reduction rates and resulted in 13.13 to 10.12% (NSICU), 13.90 to 10.98% (spine surgery), and 23.64 to 21.20% (TBI). Approximately $1,300,614.28 in saving resulted from potential interventions. CONCLUSION: This study reports the successful development and simulation of an ML-based approach for predicting and reducing 30-day hospital readmissions in neurosurgery. The intervention shows feasibility in improving patient outcomes and reducing financial losses.

Consistency of Faraday cup and ionization chamber dosimetry of proton fields and the role of nuclear interactions.

BACKGROUND: A Faraday cup (FC) facilitates a quite clean measurement of the proton fluence emerging from clinical spot-scanning nozzles with narrow pencil-beams. The utilization of FCs appears to be an attractive option for high dose rate delivery modes and the source models of Monte-Carlo (MC) dose engines. However, previous studies revealed discrepancies of 3%-6% between reference dosimetry with ionization chambers (ICs) and FC-based dosimetry. This has prevented the widespread use of FCs for dosimetry in proton therapy. PURPOSE: The current study aims at bridging the gap between FC dosimetry and IC dosimetry of proton fields delivered with spot-scanning treatment heads. Particularly, a novel method to evaluate FC measurements is introduced. METHODS: A consistency check is formulated, which makes use of the energy balance and the reciprocity theorem. The measurement data comprise central-axis depth distributions of the absorbed dose of quasi-monochromatic fields with a width of about 28.5 cm and FC measurements of the reciprocal fields with a single spot. These data are complemented by a look-up of energy-range tables, the average Q-value of transmutations, and the escape energy carried away by neutrons and photons. The latter data are computed by MC simulations, which in turn are validated with measurements of the distal dose tail and neutron out-of-field doses. For comparison, the conventional approach of FC evaluation is performed, which computes absorbed dose from the product of fluence and stopping power. The results from the FC measurements are compared with the standard dosimetry protocols and improved reference dosimetry methods. RESULTS: The deviation between the conventional FC-based dosimetry and the IC-based one according to standard dosimetry protocols was -4.7 ( ± $\pm$ 3.3)% for a 100 MeV field and -3.6 ( ± $\pm$ 3.5)% for 200 MeV, thereby agreeing within the reported uncertainties. The deviations could be reduced to -4.0 ( ± $\pm$ 2.9)% and -3.0 ( ± $\pm$ 3.1)% by adopting state-of-the-art reference dosimetry methods. The alternative approach using the energy balance gave deviations of only -1.9% (100 MeV) and -2.6% (200 MeV) using state-of-the-art dosimetry. The standard uncertainty of this novel approach was estimated to be about 2%. CONCLUSIONS: An alternative concept has been established to determine the absorbed dose of monoenergetic proton fields with an FC. It eliminates the strong dependence of the conventional FC-based approach on the MC simulation of the stopping-power and of the secondary ions, which according to the study at hand is the major contributor to the underestimation of the absorbed dose. Some contributions to the uncertainty of the novel approach could potentially be reduced in future studies. This would allow for accurate consistency tests of conventional dosimetry procedures.