Introduction Postoperative delirium in geriatric hip fracture patients adversely affects clinical and functional outcomes and increases costs. A preoperative prediction tool to identify high-risk patients may facilitate optimal use of preventive interventions. The purpose of this study was to develop a clinical prediction model using machine learning algorithms for preoperative prediction of postoperative delirium in geriatric hip fracture patients. Materials & Methods Geriatric patients undergoing operative hip fracture fixation were queried in the American College of Surgeons National Surgical Quality Improvement Program database (ACS NSQIP) from 2016 through 2019. A total of 28 207 patients were included, of which 8030 (28.5%) developed a postoperative delirium. First, the dataset was randomly split 80:20 into a training and testing subset. Then, a random forest (RF) algorithm was used to identify the variables predictive for a postoperative delirium. The machine learning-model was developed on the training set and the performance was assessed in the testing set. Performance was assessed by discrimination (c-statistic), calibration (slope and intercept), overall performance (Brier-score), and decision curve analysis. Results The included variables identified using RF algorithms were (1) age, (2) ASA class, (3) functional status, (4) preoperative dementia, (5) preoperative delirium, and (6) preoperative need for mobility-aid. The clinical prediction model reached good discrimination (c-statistic = .79), almost perfect calibration (intercept = −.01, slope = 1.02), and excellent overall model performance (Brier score = .15). The clinical prediction model was deployed as an open-access web-application: https://sorg-apps.shinyapps.io/hipfxdelirium/ . Discussion & Conclusions We developed a clinical prediction model that shows promise in estimating the risk of postoperative delirium in geriatric hip fracture patients. The clinical prediction model can play a beneficial role in decision-making for preventative measures for patients at risk of developing a delirium. If found to be externally valid, clinicians might use the available web-based application to help incorporate the model into clinical practice to aid decision-making and optimize preoperative prevention efforts.

Background: Oncologists often struggle with managing the unique care needs of older adults with cancer. This study sought to determine the feasibility of delivering a transdisciplinary intervention targeting the geriatric-specific (physical function and comorbidity) and palliative care (symptoms and prognostic understanding) needs of older adults with advanced cancer. Methods: Patients aged ≥65 years with incurable gastrointestinal or lung cancer were randomly assigned to a transdisciplinary intervention or usual care. Those in the intervention arm received 2 visits with a geriatrician, who addressed patients’ palliative care needs and conducted a geriatric assessment. We predefined the intervention as feasible if >70% of eligible patients enrolled in the study and >75% of eligible patients completed study visits and surveys. At baseline and week 12, we assessed patients’ quality of life (QoL), symptoms, and communication confidence. We calculated mean change scores in outcomes and estimated intervention effect sizes (ES; Cohen’s d) for changes from baseline to week 12, with 0.2 indicating a small effect, 0.5 a medium effect, and 0.8 a large effect. Results: From February 2017 through June 2018, we randomized 62 patients (55.9% enrollment rate [most common reason for refusal was feeling too ill]; median age, 72.3 years; cancer types: 56.5% gastrointestinal, 43.5% lung). Among intervention patients, 82.1% attended the first visit and 79.6% attended both. Overall, 89.7% completed all study surveys. Compared with usual care, intervention patients had less QoL decrement (–0.77 vs –3.84; ES = 0.21), reduced number of moderate/severe symptoms (–0.69 vs +1.04; ES = 0.58), and improved communication confidence (+1.06 vs –0.80; ES = 0.38). Conclusions: In this pilot trial, enrollment exceeded 55%, and >75% of enrollees completed all study visits and surveys. The transdisciplinary intervention targeting older patients’ unique care needs showed encouraging ES estimates for enhancing patients’ QoL, symptom burden, and communication confidence.

12012 Background: Older adults with gastrointestinal (GI) cancers undergoing surgery often experience poor outcomes, such as prolonged postoperative (post-op) length of stay (LOS), intensive care unit (ICU) use, and readmissions. Involvement of geriatricians in the care of older adults with cancer can improve outcomes. We conducted a randomized trial of a perioperative geriatric intervention in older adults with GI cancers undergoing surgery. Methods: We randomly assigned patients age ≥65 with GI cancers planning to undergo surgical resection to receive a perioperative geriatric intervention or usual care. Intervention patients met with a geriatrician preoperatively in the outpatient setting and post-op as an inpatient consultant. The geriatrician conducted a geriatric assessment and made recommendations to the surgical/oncology teams. The primary end point was post-op LOS. Secondary end points included post-op ICU use, readmission risk, and patient-reported symptom burden (Edmonton Symptom Assessment System [ESAS]) and depression symptoms (Geriatric Depression Scale). We conducted both intention-to-treat (ITT) and per protocol (PP) analyses. Results: From 9/13/16-4/30/19, we randomized 160 patients (72.4% enrollment rate; median age = 72 [65-92]). The ITT analyses included 137/160 patients who underwent surgery (usual care = 68/78, intervention = 69/82). The PP analyses included the 68 usual care patients and the 30/69 intervention patients who received both pre- and post-op intervention components. In ITT analyses, we found no significant differences between intervention and usual care in post-op LOS (7.2 v 8.2 days, P = .37), ICU use (23.3% v 32.4%, p = .23), and readmission rates within 90 days of surgery (21.7% v 25.0%, p = .65). Intervention patients reported lower depression symptoms (B = -1.39, P < .01) at post-op day 5 and fewer moderate/severe ESAS symptoms at post-op day 60 (B = -1.09, P = .02). In PP analyses, intervention patients had significantly shorter post-op LOS (5.9 v 8.2 days, P = .02) and lower rates of post-op ICU use (13.3% v 32.4%, p < .05), but readmission rates were not significantly different (16.7% v 25.0%, p = .36). Conclusions: Although this perioperative geriatric intervention did not have a significant impact on the primary end point in ITT analysis, we found encouraging results in several secondary outcomes and for the subgroup of patients who received the planned intervention. Future studies of this perioperative geriatric intervention should include efforts, such as telehealth visits, to ensure the intervention is delivered as planned. Clinical trial information: NCT02810652 .

No abstract