Cross-Combination Analyses of Random Forest Feature Selection and Decision Tree Model for Predicting Intraoperative Hypothermia in Total Joint Arthroplasty.

BACKGROUND: In total joint arthroplasty patients, intraoperative hypothermia (IOH) is associated with perioperative complications and an increased economic burden. Previous models have some limitations and mainly focus on regression modeling. Random forest (RF) algorithms and decision tree modeling are effective for eliminating irrelevant features and making predictions that aid in accelerating modeling and reducing application difficulty. METHODS: We conducted this prospective observational study using convenience sampling and collected data from 327 total joint arthroplasty patients in a tertiary hospital from March 4, 2023 to September 11, 2023. Of those, 229 patients were assigned to the training and 98 to the testing sets. The Chi-square, Mann-Whitney U, and t-tests were used for baseline analyses. The feature variables selection used the RF algorithms, and the decision tree model was trained on 299 examples and validated on 98. The sensitivity, specificity, recall, F1 score, and area under the curve (AUC) were used to test the model's performance. RESULTS: The RF algorithms identified the preheating time, the volume of flushing fluids, the intraoperative infusion volume, the anesthesia time, the surgical time, and the core temperature after intubation as risk factors for IOH. The decision tree was grown to five levels with nine terminal nodes. The overall incidence of IOH was 42.13%. The sensitivity, specificity, recall, F1 score, and AUC were 0.651, 0.907, 0.916, 0.761, and 0.810, respectively. The model indicated strong internal consistency and predictive ability. CONCLUSIONS: The preheating time, the volume of flushing fluids, the intraoperative infusion volume, the anesthesia time, the surgical time, and the core temperature after intubation could accurately predict IOH in total joint arthroplasty patients. By monitoring these factors, the clinical staff could achieve early detection and intervention of IOH in total joint arthroplasty patients.

Inadvertent perioperative hypothermia and surgical site infections after liver resection.

BACKGROUND: In the overall surgical population, inadvertent perioperative hypothermia has been associated with an increased incidence of surgical site infection (SSI). However, recent clinical trials did not validate this notion. This study aimed to investigate the potential correlation between inadvertent perioperative hypothermia and SSIs following liver resection. METHODS: This retrospective cohort study included all consecutive patients who underwent liver resection between January 2019 and December 2021 at the First Affiliated Hospital, Zhejiang University School of Medicine. Perioperative temperature managements were implemented for all patients included in the analysis. Estimated propensity score matching (PSM) was performed to reduce the baseline imbalances between the normothermia and hypothermia groups. Before and after PSM, univariate analyses were performed to evaluate the correlation between hypothermia and SSI. Multivariate regression analysis was performed to determine whether hypothermia was an independent risk factor for postoperative transfusion and major complications. Subgroup analyses were performed for diabetes mellitus, age > 65 years, and major liver resection. RESULTS: Among 4000 patients, 2206 had hypothermia (55.2%), of which 150 developed SSI (6.8%). PSM yielded 1434 individuals in each group. After PSM, the hypothermia and normothermia groups demonstrated similar incidence rates of SSI (6.3% vs. 7.0%, P = 0.453), postoperative transfusion (13.3% vs. 13.7%, P = 0.743), and major complications (9.0% vs. 10.1%, P = 0.309). Univariate regression analysis revealed no significant effects of hypothermia on the incidence of SSI in the group with the highest hypothermia exposure [odds ratio (OR) = 1.25, 95% confidence interval (CI): 0.84-1.87, P = 0.266], the group with moderate exposure (OR = 1.00, 95% CI: 0.65-1.53, P = 0.999), or the group with the lowest exposure (OR = 1.11, 95% CI: 0.73-1.65, P = 0.628). The subgroup analysis revealed similar results. Regarding liver function, patients in the hypothermia group demonstrated lower γ-glutamyl transpeptidase (37 vs. 43 U/L, P = 0.001) and alkaline phosphatase (69 vs. 72 U/L, P = 0.016). However, patients in the hypothermia group exhibited prolonged activated partial thromboplastin time (29.2 vs. 28.6 s, P < 0.01). CONCLUSIONS: In our study of patients undergoing liver resection, we found no significant association between mild perioperative hypothermia and SSI. It might be due to the perioperative temperature managements, especially active warming measures, which limited the impact of perioperative hypothermia on the occurrence of SSI.

Effect of Prewarming on Perioperative Hypothermia in Patients Undergoing Loco-Regional or General Anesthesia: A Randomized Clinical Trial.

Background and Objectives: Redistribution hypothermia occurs during anesthesia despite active intraoperative warming. Prewarming increases the heat absorption by peripheral tissue, reducing the central to peripheral heat gradient. Therefore, the addition of prewarming may offer a greater preservation of intraoperative normothermia as compared to intraoperative warming only. Materials and Methods: A single-center clinical trial of adults scheduled for non-cardiac surgery. Patients were randomized to receive or not a prewarming period (at least 10 min) with convective air devices. Intraoperative temperature management was identical in both groups and performed according to a local protocol. The primary endpoint was the incidence, the magnitude and the duration of hypothermia (according to surgical time) between anesthetic induction and arrival at the recovery room. Secondary outcomes were core temperature on arrival in operating room, surgical site infections, blood losses, transfusions, patient discomfort (i.e., shivering), reintervention and hospital stay. Results: In total, 197 patients were analyzed: 104 in the control group and 93 in the prewarming group. Core temperature during the intra-operative period was similar between groups (p = 0.45). Median prewarming lasted 27 (17-38) min. Regarding hypothermia, we found no differences in incidence (controls: 33.7%, prewarming: 39.8%; p = 0.37), duration (controls: 41.6% (17.8-78.1), prewarming: 45.2% (20.6-71.1); p = 0.83) and magnitude (controls: 0.19 °C · h-1 (0.09-0.54), prewarming: 0.20 °C · h-1 (0.05-0.70); p = 0.91). Preoperative thermal discomfort was more frequent in the prewarming group (15.1% vs. 0%; p < 0.01). The interruption of intraoperative warming was more common in the prewarming group (16.1% vs. 6.7%; p = 0.03), but no differences were seen in other secondary endpoints. Conclusions: A preoperative prewarming period does not reduce the incidence, duration and magnitude of intraoperative hypothermia. These results should be interpreted considering a strict protocol for perioperative temperature management and the low incidence of hypothermia in controls.

[Correlation Between 6 Characteristics of Perioperative Hypothermia and Allogeneic Red Blood Cell Transfusion in Abdominal Surgery Patients]. / è ¹éƒ¨æ‰‹æœ¯æ‚£è€ å›´æœ¯æœŸå ­ç§ä½Žä½“æ¸©ç‰¹å¾ä¸Žå¼‚ä½“çº¢ç»†èƒžè¾“æ³¨çš„å ³è”æ€§åˆ†æž.

Objective: To explore the correlation between six characteristics of perioperative hypothermia and allogeneic red blood cell (RBC) transfusions in patients who underwent abdominal surgeries. Methods: Patients who underwent abdominal surgeries at West China Hospital, Sichuan University between October 2019 and July 2021 were retrospectively enrolled. A wearable wireless temperature sensor was used to continuously monitor the core body temperature of patients throughout the perioperative period. The perioperative temperature nadir, maximum temperature loss, percentage of time with hypothermia, time-weighted average temperature, area under the curve (AUC) at 36 ℃, and AUC at 37 ℃ were calculated for the period from entering the operation room to 24 hours after the end of anesthesia. The restricted cubic spline (RCS) and multiple logistic regression models were used to explore the correlation between these temperature characteristics and perioperative allogeneic RBC transfusions. Results: A total of 3119 patients were included in the study, with an allogeneic RBC transfusion rate of 2.8%. The RCS model showed that allogeneic RBC transfusion was associated with the perioperative temperature nadir (Poverall=0.048) and AUC at 36 ℃ (Poverall=0.026) and no statistical significance was found in the nonlinear test. The association between allogeneic RBC transfusions and other temperature characteristics was not statistically significant. According to the RCS model results, cut-off points were taken to form groups based on the body temperature characteristics. Multivariate logistic regression showed that the perioperative temperature nadir<35.5 ℃ (odds ratio [OR]=2.47, 95% confidence interval [CI]: 1.21-5.03) and AUC at 36 ℃≥100 ℃·min (OR=2.24, 95% CI:1.09-4.58) were associated with increased demand for allogeneic RBC transfusion. Conclusion: Hypothermia is associated with an increased need for perioperative allogeneic RBC transfusions and has a cumulative effect over time. For patients at high risk of bleeding, attention should be paid to the prevention of perioperative hypothermia and reduction in the cumulative exposure to hypothermia, thereby reducing the need for blood transfusion.

Inadvertent perioperative hypothermia prevention strategies for urology surgical patients who received a blood transfusion: A retrospective analysis.

OBJECTIVES: This study aimed to establish whether hypothermia was present in patients who required a blood transfusion and underwent a urology procedure, as well as identify staff knowledge and understanding. PATIENTS AND METHODS: A staff survey was conducted with respondents from a range of clinical settings, with some staff working across more than one area. A retrospective review of 46 medical records was conducted between January 2021 and July 2022. All data were exported into an Excel spreadsheet and analysed. RESULTS: Staff (70%) were unaware of guidelines informing thermoregulation practices; however, 90% understood the importance of normothermia in the perioperative environment. Medical record review demonstrated temperature monitoring and intervention implementation varied across the perioperative journey, with 20% of patients hypothermic on admission and 89% of the cohort having two or more risk factors. CONCLUSION: There is no formal process for the management of inadvertent perioperative hypothermia throughout the patient journey at the hospital. A variety of intrinsic factors (age, patient comorbidities, American Society of Anaesthesiologists score) and external factors (patient waiting times, anaesthetic modality, type of procedure, environmental influences), impact each patient's risk of inadvertent perioperative hypothermia.