20Background: We investigated the incidence of acute kidney injury (AKI) and the risk 21 factors associated with vancomycin (VAN) and piperacillin-tazobactam (TZP) 22 combination therapy in non-intensive care unit (ICU) and ICU settings. 23Methods: In this single-center retrospective cohort study, adult patients who primary end point was the incidence of AKI. 26Results: A total of 593 adult patients were included. The incidence of AKI was 10.6% 27 overall, 8.0% in the non-TZP group, and 19.8% in the TZP group. In a univariate 28 analysis, the odds of AKI were higher in the TZP group than in the non-TZP group 29 (odds ratio (OR) = 2.84, 95% confidence interval (CI) = 1. 64-4.90). In both non-ICU 30 and ICU settings, the odds of AKI were higher in the TZP group than in the non-TZP 31 group (non-ICU: OR = 3.04, 95% CI = 1.52-6.09; ICU: OR = 2.51, 95% CI = 32 1.03-6.08). Furthermore, a propensity score analysis revealed that the odds of AKI were 33 higher in the TZP group than in the non-TZP group (OR = 2.81, 95% CI = 1.52-5.17). 34In both non-ICU and ICU settings, the odds of AKI were higher in the TZP group than 35 in the non-TZP group (non-ICU: OR = 2.57, 95% CI = 1.17-5.64; ICU: OR = 3.51, 95% 36 CI = 1.05-11.6). 37 3 Conclusions: The combined use of TZP in patients receiving VAN increased the 38 incidence of AKI in both non-ICU and ICU settings.39 40 Keywords: Acute kidney injury, Vancomycin, Piperacillin-tazobactam, Intensive care 41 unit, Nephrotoxicity 42 43 44 48 of gram-positive pathogens, including methicillin-resistant Staphylococcus aureus. The 49risk of developing AKI during VAN treatment is widely known and its incidence rate is 50 reported to be approximately 5 to 7% 3,4 . High-dose administration and elevated trough 51 concentrations in VAN treatment are associated with AKI 5,6 . 52In recent years, multiple studies have suggested an association between AKI and 53 VAN plus piperacillin-tazobactam (TZP) combination therapy 7-14 . However, past 54 reports about the risk of developing AKI after VAN and TZP combination therapy are 55 small observational studies with varying results. In addition, the combination of drugs 56 that is associated with increased AKI risk in an intensive care unit (ICU) setting is 57 controversial 9-12 . 58Therefore, the purpose of this study was to determine the incidence of AKI and 59 the risk factors associated with VAN and TZP combination therapy in non-ICU and 60 ICU settings. We considered AKI by dividing the non-ICU group and the ICU group, 61 5 assessed whether the incidence of AKI due to VAN plus TZP differs depending on the 62 severity of patients. 63 64 Materials and methods 65 Study design and setting 66 This single-center, retrospective cohort study was conducted at Nippon Medical School 67 Hospital, a tertiary academic hospital in Tokyo, Japan. This study was reviewed and 68 approved by the Institutional Review Board of Nippon Medical School Hospital before 69 its conduct (30-05-937).70 71 Study population 72 2017 at Nippon Medical School were included. Pa...

Background: It is difficult to predict vancomycin trough concentrations in critically ill patients as their pharmacokinetics change with the progression of both organ failure and medical intervention. This study aims to develop a model to predict vancomycin trough concentration using machine learning (ML) and to compare its prediction accuracy with that of the population pharmacokinetic (PPK) model. Methods: A single-center retrospective observational study was conducted. Patients who had been admitted to the intensive care unit, received intravenous vancomycin, and had undergone therapeutic drug monitoring between 2013 and 2020,were included. Thereafter, ML models were developed with random forest, LightGBM, and ridge regression using 42 features. Mean absolute errors (MAE) were compared and important features were shown using LightGBM. Results: Among 335 patients, 225 were included as training data and 110 were used for test data. A significant difference was identified in the MAE by each ML model compared with PPK;4.13 ± 3.64 for random forest, 4.18 ± 3.37 for LightGBM, 4.29 ± 3.88 for ridge regression, and 6.17 ± 5.36 for PPK. The highest importance features were pH, lactate, and serum creatinine. Conclusion: This study concludes that ML may be able to more accurately predict vancomycin trough concentrations than the currently used PPK model in ICU patients.