Multi-view representation learning for tabular data integration using inter-feature relationships.

OBJECTIVE: An applied problem facing all areas of data science is harmonizing data sources. Joining data from multiple origins with unmapped and only partially overlapping features is a prerequisite to developing and testing robust, generalizable algorithms, especially in healthcare. This integrating is usually resolved using meta-data such as feature names, which may be unavailable or ambiguous. Our goal is to design methods that create a mapping between structured tabular datasets derived from electronic health records independent of meta-data. METHODS: We evaluate methods in the challenging case of numeric features without reliable and distinctive univariate summaries, such as nearly Gaussian and binary features. We assume that a small set of features are a priori mapped between two datasets, which share unknown identical features and possibly many unrelated features. Inter-feature relationships are the main source of identification which we expect. We compare the performance of contrastive learning methods for feature representations, novel partial auto-encoders, mutual-information graph optimizers, and simple statistical baselines on simulated data, public datasets, the MIMIC-III medical-record changeover, and perioperative records from before and after a medical-record system change. Performance was evaluated using both mapping of identical features and reconstruction accuracy of examples in the format of the other dataset. RESULTS: Contrastive learning-based methods overall performed the best, often substantially beating the literature baseline in matching and reconstruction, especially in the more challenging real data experiments. Partial auto-encoder methods showed on-par matching with contrastive methods in all synthetic and some real datasets, along with good reconstruction. However, the statistical method we created performed reasonably well in many cases, with much less dependence on hyperparameter tuning. When validating feature match output in the EHR dataset we found that some mistakes were actually a surrogate or related feature as reviewed by two subject matter experts. CONCLUSION: In simulation studies and real-world examples, we find that inter-feature relationships are effective at identifying matching or closely related features across tabular datasets when meta-data is not available. Decoder architectures are also reasonably effective at imputing features without an exact match.

Identifying the blue patient: preoperative screening for depression.

Preoperative depression is an underappreciated comorbidity that has important implications for postoperative outcomes. Screening for symptoms of depression before surgery can identify patients with or without a previous diagnosis of depression who could benefit from perioperative interventions to improve mood. Preoperative screening programmes are feasible to implement, although care must be taken to ensure that patients who are most likely to benefit are included.

Preoperative predictors of acute postoperative anxiety and depression using ecological momentary assessments: a secondary analysis of a single-centre prospective observational study.

BACKGROUND: Postoperative anxiety and depression can negatively affect surgical outcomes and patient wellbeing. This study aimed to quantify the incidence of postoperative worsening anxiety and depression symptoms and to identify preoperative predictors of these conditions. METHODS: This prospective, observational cohort study included 1168 patients undergoing surgery lasting >1 h with overnight admission at a university-affiliated quaternary referral centre. Postoperative anxiety and depression symptoms were measured using standardised, thrice-daily ecological momentary assessments (EMAs) for 30 days. Co-primary outcomes were worsening anxiety and depression symptoms, each defined as a slope >0 when EMA was modelled as a linear function of time. Multivariable logistic regression was performed to identify independent preoperative predictors of each outcome. RESULTS: Postoperative worsening anxiety occurred in 60 patients (5%), and postoperative worsening depression occurred in 86 patients (7%). Predictors of postoperative worsening of anxiety symptoms included preoperative Patient-Reported Outcome Measurement Information System (PROMIS) anxiety symptoms (adjusted odds ratio [aOR] 2.48, 95% credible interval [CI] 1.29-4.79, for mild symptoms; aOR 2.22, 95% CI 1.10-4.51, for moderate to severe symptoms), and preoperative pain (aOR 3.46, 95% CI 1.32-9.12). Predictors of postoperative worsening depression symptoms included preoperative PROMIS depression symptoms (aOR 2.26, 95% CI 1.24-4.14, for mild symptoms; aOR 3.79, 95% CI 2.10-6.81, for moderate to severe symptoms). Self-reported history of anxiety or depression did not independently predict either outcome. CONCLUSIONS: Postoperative worsening anxiety and depression appear to be associated more closely with preoperative active mental health or pain symptoms rather than self-reported history of these conditions. Preoperative identification of at-risk patients will require screening for symptoms rather than simple history taking.

Effect of machine learning models on clinician prediction of postoperative complications: the Perioperative ORACLE randomised clinical trial.

BACKGROUND: Anaesthesiologists might be able to mitigate risk if they know which patients are at greatest risk for postoperative complications. This trial examined the impact of machine learning models on clinician risk assessment. METHODS: This single-centre, prospective, randomised clinical trial enrolled surgical patients aged ≥18 yr. Anaesthesiologists and nurse anaesthetists providing remote telemedicine support reviewed electronic health records with (assisted group) or without (unassisted group) reviewing machine learning predictions. Clinicians predicted the likelihood of postoperative 30-day all-cause mortality and postoperative acute kidney injury (AKI) within 7 days. The primary outcome was area under the receiver operating characteristic curve (AUROC) for clinician predictions of mortality and AKI, comparing AUROCs between assisted and unassisted assessments. RESULTS: We analysed 5071 patients (mean [range] age: 58 [18-100] yr; 52% female) assessed by 89 clinicians. Of these, 98 (2.2%) patients died within 30 days of surgery and 450 (11.1%) patients sustained AKI. Clinician predictions agreed with the models more strongly in the assisted vs unassisted group (weighted kappa 0.75 vs 0.62 for death, mean difference: 0.13 [95% CI 0.10-0.17]; and 0.79 vs 0.54 for AKI, mean difference: 0.25 [95% CI 0.21-0.29]). Clinical prediction of death was similar between the assisted (AUROC 0.793) and unassisted (AUROC 0.780) groups (mean difference: 0.013 [95% CI -0.070 to 0.097]; P=0.76). Prediction of AKI had an AUROC of 0.734 in the assisted group vs 0.688 in the unassisted group (difference 0.046 [95% CI -0.003 to 0.091]; P=0.06). CONCLUSIONS: Clinician performance was not improved by machine learning assistance. Further work is needed to clarify the role of machine learning in real-time perioperative risk stratification. CLINICAL TRIAL REGISTRATION: NCT05042804.

User-Centered Design of a Machine Learning Dashboard for Prediction of Postoperative Complications.

BACKGROUND: Machine learning models can help anesthesiology clinicians assess patients and make clinical and operational decisions, but well-designed human-computer interfaces are necessary for machine learning model predictions to result in clinician actions that help patients. Therefore, the goal of this study was to apply a user-centered design framework to create a user interface for displaying machine learning model predictions of postoperative complications to anesthesiology clinicians. METHODS: Twenty-five anesthesiology clinicians (attending anesthesiologists, resident physicians, and certified registered nurse anesthetists) participated in a 3-phase study that included (phase 1) semistructured focus group interviews and a card sorting activity to characterize user workflows and needs; (phase 2) simulated patient evaluation incorporating a low-fidelity static prototype display interface followed by a semistructured interview; and (phase 3) simulated patient evaluation with concurrent think-aloud incorporating a high-fidelity prototype display interface in the electronic health record. In each phase, data analysis included open coding of session transcripts and thematic analysis. RESULTS: During the needs assessment phase (phase 1), participants voiced that (a) identifying preventable risk related to modifiable risk factors is more important than nonpreventable risk, (b) comprehensive patient evaluation follows a systematic approach that relies heavily on the electronic health record, and (c) an easy-to-use display interface should have a simple layout that uses color and graphs to minimize time and energy spent reading it. When performing simulations using the low-fidelity prototype (phase 2), participants reported that (a) the machine learning predictions helped them to evaluate patient risk, (b) additional information about how to act on the risk estimate would be useful, and (c) correctable problems related to textual content existed. When performing simulations using the high-fidelity prototype (phase 3), usability problems predominantly related to the presentation of information and functionality. Despite the usability problems, participants rated the system highly on the System Usability Scale (mean score, 82.5; standard deviation, 10.5). CONCLUSIONS: Incorporating user needs and preferences into the design of a machine learning dashboard results in a display interface that clinicians rate as highly usable. Because the system demonstrates usability, evaluation of the effects of implementation on both process and clinical outcomes is warranted.

Brain-Hazardous Medications and Potential Subadequate Antidepressant Dosing in Older Surgical Patients Receiving Home Antidepressants: An Observational Study of a Large US Health System.

BACKGROUND: Older surgical patients with depression often experience poor postoperative outcomes. Poor outcomes may stem from brain-hazardous medications and subadequate antidepressant dosing. METHODS: This was a retrospective, observational cohort study covering the period between January 1, 2021 and December 31, 2021. Patients ≥60 years of age who underwent inpatient surgery and had an overnight stay at an integrated academic health care system comprising 14 hospitals were eligible. We analyzed the prevalence of home central nervous system (CNS)-active potentially inappropriate medication (PIM) and potential subadequate antidepressant dosing in older surgical patients receiving home antidepressants. Univariable and multivariable regression models were used to identify factors associated with home CNS-active PIM prescribing and potential subadequate antidepressant dosing. Additionally, outcomes were compared among patients receiving and not receiving CNS-active PIMs and patients receiving and not receiving subadequate antidepressant dosing. RESULTS: A total of 8031 patients were included in this study (47% female, mean age = 70 years) of whom 2087 (26%) were prescribed antidepressants. Roughly one-half (49%, 95% confidence interval [CI], 46.5-50.1) of patients receiving home antidepressants were also receiving ≥1 CNS-active PIM and 29% (95% CI, 27.0-29.3) were receiving a potential subadequate dose. Factors associated with an increased likelihood of receiving a home CNS-active PIM included female sex (adjusted odds ratio [aOR], 1.46), anxiety (aOR, 2.43), asthma or chronic obstructive pulmonary disease (aOR, 1.39), and serotonin-norepinephrine reuptake inhibitor use (aOR, 1.54). Patients aged ≥75 years (aOR, 1.57), black race (aOR, 1.48) and those with congestive heart failure (aOR, 1.33) were more likely to be prescribed a potential subadequate antidepressant dose. Patients receiving potential subadequate antidepressant doses were discharged home less often (64% vs 73%), had a longer hospital length of stay (9 days vs 7 days), and a higher mortality rate (18% vs 10%) compared to patients receiving adequate home antidepressant doses (P-value for all <0.01). No differences in these outcomes were found among patients receiving home antidepressants with or without CNS-active PIMs. CONCLUSIONS: Older surgical patients receiving antidepressants are frequently prescribed brain-hazardous medications and potentially subadequate antidepressant doses. Those receiving subadequate antidepressant doses may be at risk for worse postoperative outcomes compared to patients receiving adequate doses. The role of preoperative medication optimization to improve outcomes for older surgical patients should be evaluated.

Electroencephalography-Guided Anesthesia and Delirium in Older Adults After Cardiac Surgery: The ENGAGES-Canada Randomized Clinical Trial.

Importance: Intraoperative electroencephalogram (EEG) waveform suppression, suggesting excessive general anesthesia, has been associated with postoperative delirium. Objective: To assess whether EEG-guided anesthesia decreases the incidence of delirium after cardiac surgery. Design, Setting, and Participants: Randomized, parallel-group clinical trial of 1140 adults 60 years or older undergoing cardiac surgery at 4 Canadian hospitals. Recruitment was from December 2016 to February 2022, with follow-up until February 2023. Interventions: Patients were randomized in a 1:1 ratio (stratified by hospital) to receive EEG-guided anesthesia (n = 567) or usual care (n = 573). Patients and those assessing outcomes were blinded to group assignment. Main Outcomes and Measures: The primary outcome was delirium during postoperative days 1 through 5. Intraoperative measures included anesthetic concentration and EEG suppression time. Secondary outcomes included intensive care and hospital length of stay. Serious adverse events included intraoperative awareness, medical complications, and 30-day mortality. Results: Of 1140 randomized patients (median [IQR] age, 70 [65-75] years; 282 [24.7%] women), 1131 (99.2%) were assessed for the primary outcome. Delirium during postoperative days 1 to 5 occurred in 102 of 562 patients (18.15%) in the EEG-guided group and 103 of 569 patients (18.10%) in the usual care group (difference, 0.05% [95% CI, -4.57% to 4.67%]). In the EEG-guided group compared with the usual care group, the median volatile anesthetic minimum alveolar concentration was 0.14 (95% CI, 0.15 to 0.13) lower (0.66 vs 0.80) and there was a 7.7-minute (95% CI, 10.6 to 4.7) decrease in the median total time spent with EEG suppression (4.0 vs 11.7 min). There were no significant differences between groups in median length of intensive care unit (difference, 0 days [95% CI, -0.31 to 0.31]) or hospital stay (difference, 0 days [95% CI, -0.94 to 0.94]). No patients reported intraoperative awareness. Medical complications occurred in 64 of 567 patients (11.3%) in the EEG-guided group and 73 of 573 (12.7%) in the usual care group. Thirty-day mortality occurred in 8 of 567 patients (1.4%) in the EEG-guided group and 13 of 573 (2.3%) in the usual care group. Conclusions and Relevance: Among older adults undergoing cardiac surgery, EEG-guided anesthetic administration to minimize EEG suppression, compared with usual care, did not decrease the incidence of postoperative delirium. This finding does not support EEG-guided anesthesia for this indication. Trial Registration: ClinicalTrials.gov Identifier: NCT02692300.

Pro-Con Debate: Electroencephalography-Guided Anesthesia for Reducing Postoperative Delirium.

Postoperative delirium (POD) has significant implications on morbidity, mortality, and health care expenditures. Monitoring electroencephalography (EEG) to adjust anesthetic management has gained interest as a strategy to mitigate POD. In this Pro-Con commentary article, the pro side supports the use of EEG to reduce POD, citing an empiric reduction in POD with processed EEG (pEEG)-guided general anesthesia found in several studies and recent meta-analysis. The Electroencephalography Guidance of Anesthesia to Alleviate Geriatric Syndromes (ENGAGES) trial is the exception to this, and issues with methods and achieved depths are discussed. Meanwhile, the Con side advocates that the use of EEG to reduce POD is not yet certain, citing that there is a lack of evidence that associations between anesthetic depth and POD represent causal relationships. The Con side also contends that the ideal EEG signatures to guide anesthetic titration are currently unknown, and the potential benefits of reduced anesthesia levels may be outweighed by the risks of potentially insufficient anesthetic administration. As the public health burden of POD increases, anesthesia clinicians will be tasked to consider interventions to mitigate risk such as EEG. This Pro-Con debate will provide 2 perspectives on the evidence and rationales for using EEG to mitigate POD.

Association of Race and Neighborhood Disadvantage with Patient Engagement in a Home-Based COVID-19 Remote Monitoring Program.

BACKGROUND: COVID-positive outpatients may benefit from remote monitoring, but such a program often relies on smartphone apps. This may introduce racial and socio-economic barriers to participation. Offering multiple methods for participation may address these barriers. OBJECTIVES: (1) To examine associations of race and neighborhood disadvantage with patient retention in a monitoring program offering two participation methods. (2) To measure the association of the program with emergency department visits and hospital admissions. DESIGN: Retrospective propensity-matched cohort study. PARTICIPANTS: COVID-positive outpatients at a single university-affiliated healthcare system and propensity-matched controls. INTERVENTIONS: A home monitoring program providing daily symptom tracking via patient portal app or telephone calls. MAIN MEASURES: Among program enrollees, retention (until 14 days, symptom resolution, or hospital admission) by race and neighborhood disadvantage, with stratification by program arm. In enrollees versus matched controls, emergency department utilization and hospital admission within 30 days. KEY RESULTS: There were 7592 enrolled patients and 9710 matched controls. Black enrollees chose the telephone arm more frequently than White enrollees (68% versus 44%, p = 0.009), as did those from more versus less disadvantaged neighborhoods (59% versus 43%, p = 0.02). Retention was similar in Black enrollees and White enrollees (63% versus 62%, p = 0.76) and in more versus less disadvantaged neighborhoods (63% versus 62%, p = 0.44). When stratified by program arm, Black enrollees had lower retention than White enrollees in the app arm (49% versus 55%, p = 0.01), but not in the telephone arm (69% versus 71%, p = 0.12). Compared to controls, enrollees more frequently visited the emergency department (HR 1.71 [95% CI 1.56-1.87]) and were admitted to the hospital (HR 1.16 [95% CI 1.02-1.31]). CONCLUSIONS: In a COVID-19 remote patient monitoring program, Black enrollees preferentially selected, and had higher retention in, telephone- over app-based monitoring. As a result, overall retention was similar between races. Remote monitoring programs with multiple modes may reduce barriers to participation.

Effect of electroencephalogram-guided anaesthesia administration on 1-yr mortality: follow-up of a randomised clinical trial.

BACKGROUND: Intraoperative EEG suppression duration has been associated with postoperative delirium and mortality. In a clinical trial testing anaesthesia titration to avoid EEG suppression, the intervention did not decrease the incidence of postoperative delirium, but was associated with reduced 30-day mortality. The present study evaluated whether the EEG-guided anaesthesia intervention was also associated with reduced 1-yr mortality. METHODS: This manuscript reports 1 yr follow-up of subjects from a single-centre RCT, including a post hoc secondary outcome (1-yr mortality) in addition to pre-specified secondary outcomes. The trial included subjects aged 60 yr or older undergoing surgery with general anaesthesia between January 2015 and May 2018. Patients were randomised to receive EEG-guided anaesthesia or usual care. The previously reported primary outcome was postoperative delirium. The outcome of the current study was all-cause 1-yr mortality. RESULTS: Of the 1232 subjects enrolled, 614 subjects were randomised to EEG-guided anaesthesia and 618 subjects to usual care. One-year mortality was 57/591 (9.6%) in the guided group and 62/601 (10.3%) in the usual-care group. No significant difference in mortality was observed (adjusted absolute risk difference, -0.7%; 99.5% confidence interval, -5.8% to 4.3%; P=0.68). CONCLUSIONS: An EEG-guided anaesthesia intervention aiming to decrease duration of EEG suppression during surgery did not significantly decrease 1-yr mortality. These findings, in the context of other studies, do not provide supportive evidence for EEG-guided anaesthesia to prevent intermediate term postoperative death. CLINICAL TRIAL REGISTRATION: NCT02241655.

Preoperative Cognitive Abnormality, Intraoperative Electroencephalogram Suppression, and Postoperative Delirium: A Mediation Analysis.

BACKGROUND: Postoperative delirium is a common complication that hinders recovery after surgery. Intraoperative electroencephalogram suppression has been linked to postoperative delirium, but it is unknown if this relationship is causal or if electroencephalogram suppression is merely a marker of underlying cognitive abnormalities. The hypothesis of this study was that intraoperative electroencephalogram suppression mediates a nonzero portion of the effect between preoperative abnormal cognition and postoperative delirium. METHODS: This is a prespecified secondary analysis of the Electroencephalography Guidance of Anesthesia to Alleviate Geriatric Syndromes (ENGAGES) randomized trial, which enrolled patients age 60 yr or older undergoing surgery with general anesthesia at a single academic medical center between January 2015 and May 2018. Patients were randomized to electroencephalogram-guided anesthesia or usual care. Preoperative abnormal cognition was defined as a composite of previous delirium, Short Blessed Test cognitive score greater than 4 points, or Eight Item Interview to Differentiate Aging and Dementia score greater than 1 point. Duration of intraoperative electroencephalogram suppression was defined as number of minutes with suppression ratio greater than 1%. Postoperative delirium was detected via Confusion Assessment Method or chart review on postoperative days 1 to 5. RESULTS: Among 1,113 patients, 430 patients showed evidence of preoperative abnormal cognition. These patients had an increased incidence of postoperative delirium (151 of 430 [35%] vs.123 of 683 [18%], P < 0.001). Of this 17.2% total effect size (99.5% CI, 9.3 to 25.1%), an absolute 2.4% (99.5% CI, 0.6 to 4.8%) was an indirect effect mediated by electroencephalogram suppression, while an absolute 14.8% (99.5% CI, 7.2 to 22.5%) was a direct effect of preoperative abnormal cognition. Randomization to electroencephalogram-guided anesthesia did not change the mediated effect size (P = 0.078 for moderation). CONCLUSIONS: A small portion of the total effect of preoperative abnormal cognition on postoperative delirium was mediated by electroencephalogram suppression. Study precision was too low to determine if the intervention changed the mediated effect.

Deep-learning model for predicting 30-day postoperative mortality.

BACKGROUND: Postoperative mortality occurs in 1-2% of patients undergoing major inpatient surgery. The currently available prediction tools using summaries of intraoperative data are limited by their inability to reflect shifting risk associated with intraoperative physiological perturbations. We sought to compare similar benchmarks to a deep-learning algorithm predicting postoperative 30-day mortality. METHODS: We constructed a multipath convolutional neural network model using patient characteristics, co-morbid conditions, preoperative laboratory values, and intraoperative numerical data from patients undergoing surgery with tracheal intubation at a single medical centre. Data for 60 min prior to a randomly selected time point were utilised. Model performance was compared with a deep neural network, a random forest, a support vector machine, and a logistic regression using predetermined summary statistics of intraoperative data. RESULTS: Of 95 907 patients, 941 (1%) died within 30 days. The multipath convolutional neural network predicted postoperative 30-day mortality with an area under the receiver operating characteristic curve of 0.867 (95% confidence interval [CI]: 0.835-0.899). This was higher than that for the deep neural network (0.825; 95% CI: 0.790-0.860), random forest (0.848; 95% CI: 0.815-0.882), support vector machine (0.836; 95% CI: 0.802-870), and logistic regression (0.837; 95% CI: 0.803-0.871). CONCLUSIONS: A deep-learning time-series model improves prediction compared with models with simple summaries of intraoperative data. We have created a model that can be used in real time to detect dynamic changes in a patient's risk for postoperative mortality.

Effect of Electroencephalography-Guided Anesthetic Administration on Postoperative Delirium Among Older Adults Undergoing Major Surgery: The ENGAGES Randomized Clinical Trial.

Importance: Intraoperative electroencephalogram (EEG) waveform suppression, often suggesting excessive general anesthesia, has been associated with postoperative delirium. Objective: To assess whether EEG-guided anesthetic administration decreases the incidence of postoperative delirium. Design, Setting, and Participants: Randomized clinical trial of 1232 adults aged 60 years and older undergoing major surgery and receiving general anesthesia at Barnes-Jewish Hospital in St Louis. Recruitment was from January 2015 to May 2018, with follow-up until July 2018. Interventions: Patients were randomized 1:1 (stratified by cardiac vs noncardiac surgery and positive vs negative recent fall history) to receive EEG-guided anesthetic administration (n = 614) or usual anesthetic care (n = 618). Main Outcomes and Measures: The primary outcome was incident delirium during postoperative days 1 through 5. Intraoperative measures included anesthetic concentration, EEG suppression, and hypotension. Adverse events included undesirable intraoperative movement, intraoperative awareness with recall, postoperative nausea and vomiting, medical complications, and death. Results: Of the 1232 randomized patients (median age, 69 years [range, 60 to 95]; 563 women [45.7%]), 1213 (98.5%) were assessed for the primary outcome. Delirium during postoperative days 1 to 5 occurred in 157 of 604 patients (26.0%) in the guided group and 140 of 609 patients (23.0%) in the usual care group (difference, 3.0% [95% CI, -2.0% to 8.0%]; P = .22). Median end-tidal volatile anesthetic concentration was significantly lower in the guided group than the usual care group (0.69 vs 0.80 minimum alveolar concentration; difference, -0.11 [95% CI, -0.13 to -0.10), and median cumulative time with EEG suppression was significantly less (7 vs 13 minutes; difference, -6.0 [95% CI, -9.9 to -2.1]). There was no significant difference between groups in the median cumulative time with mean arterial pressure below 60 mm Hg (7 vs 7 minutes; difference, 0.0 [95% CI, -1.7 to 1.7]). Undesirable movement occurred in 137 patients (22.3%) in the guided and 95 (15.4%) in the usual care group. No patients reported intraoperative awareness. Postoperative nausea and vomiting was reported in 48 patients (7.8%) in the guided and 55 patients (8.9%) in the usual care group. Serious adverse events were reported in 124 patients (20.2%) in the guided and 130 (21.0%) in the usual care group. Within 30 days of surgery, 4 patients (0.65%) in the guided group and 19 (3.07%) in the usual care group died. Conclusions and Relevance: Among older adults undergoing major surgery, EEG-guided anesthetic administration, compared with usual care, did not decrease the incidence of postoperative delirium. This finding does not support the use of EEG-guided anesthetic administration for this indication. Trial Registration: ClinicalTrials.gov Identifier: NCT02241655.

Intraoperative ketamine for prevention of postoperative delirium or pain after major surgery in older adults: an international, multicentre, double-blind, randomised clinical trial.

BACKGROUND: Delirium is a common and serious postoperative complication. Subanaesthetic ketamine is often administered intraoperatively for postoperative analgesia, and some evidence suggests that ketamine prevents delirium. The primary purpose of this trial was to assess the effectiveness of ketamine for prevention of postoperative delirium in older adults. METHODS: The Prevention of Delirium and Complications Associated with Surgical Treatments [PODCAST] study is a multicentre, international randomised trial that enrolled adults older than 60 years undergoing major cardiac and non-cardiac surgery under general anaesthesia. Using a computer-generated randomisation sequence we randomly assigned patients to one of three groups in blocks of 15 to receive placebo (normal saline), low-dose ketamine (0·5 mg/kg), or high dose ketamine (1·0 mg/kg) after induction of anaesthesia, before surgical incision. Participants, clinicians, and investigators were blinded to group assignment. Delirium was assessed twice daily in the first 3 postoperative days using the Confusion Assessment Method. We did analyses by intention-to-treat and assessed adverse events. This trial is registered with clinicaltrials.gov, number NCT01690988. FINDINGS: Between Feb 6, 2014, and June 26, 2016, 1360 patients were assessed, and 672 were randomly assigned, with 222 in the placebo group, 227 in the 0·5 mg/kg ketamine group, and 223 in the 1·0 mg/kg ketamine group. There was no difference in delirium incidence between patients in the combined ketamine groups and the placebo group (19·45% vs 19·82%, respectively; absolute difference 0·36%, 95% CI -6·07 to 7·38, p=0·92). There were more postoperative hallucinations (p=0·01) and nightmares (p=0·03) with increasing ketamine doses compared with placebo. Adverse events (cardiovascular, renal, infectious, gastrointestinal, and bleeding), whether viewed individually (p value for each >0·40) or collectively (36·9% in placebo, 39·6% in 0·5 mg/kg ketamine, and 40·8% in 1·0 mg/kg ketamine groups, p=0·69), did not differ significantly across groups. INTERPRETATION: A single subanaesthetic dose of ketamine did not decrease delirium in older adults after major surgery, and might cause harm by inducing negative experiences. FUNDING: National Institutes of Health and Cancer Center Support.

Convergent Validity of Three Methods for Measuring Postoperative Complications.

BACKGROUND: Anesthesiologists need tools to accurately track postoperative outcomes. The accuracy of patient report in identifying a wide variety of postoperative complications after diverse surgical procedures has not previously been investigated. METHODS: In this cohort study, 1,578 adult surgical patients completed a survey at least 30 days after their procedure asking if they had experienced any of 18 complications while in the hospital after surgery. Patient responses were compared to the results of an automated electronic chart review and (for a random subset of 750 patients) to a manual chart review. Results from automated chart review were also compared to those from manual chart review. Forty-two randomly selected patients were contacted by telephone to explore reasons for discrepancies between patient report and manual chart review. RESULTS: Comparisons between patient report, automated chart review, and manual chart review demonstrated poor-to-moderate positive agreement (range, 0 to 58%) and excellent negative agreement (range, 82 to 100%). Discordance between patient report and manual chart review was frequently explicable by patients reporting events that happened outside the time period of interest. CONCLUSIONS: Patient report can provide information about subjective experiences or events that happen after hospital discharge, but often yields different results from chart review for specific in-hospital complications. Effective in-hospital communication with patients and thoughtful survey design may increase the quality of patient-reported complication data.

Intraoperative Electroencephalogram Suppression Predicts Postoperative Delirium.

BACKGROUND: Postoperative delirium is a common complication associated with increased morbidity and mortality, longer hospital stays, and greater health care expenditures. Intraoperative electroencephalogram (EEG) slowing has been associated previously with postoperative delirium, but the relationship between intraoperative EEG suppression and postoperative delirium has not been investigated. METHODS: In this observational cohort study, 727 adult patients who received general anesthesia with planned intensive care unit admission were included. Duration of intraoperative EEG suppression was recorded from a frontal EEG channel (FP1 to F7). Delirium was assessed twice daily on postoperative days 1 through 5 with the Confusion Assessment Method for the intensive care unit. Thirty days after surgery, quality of life, functional independence, and cognitive ability were measured using the Veterans RAND 12-item survey, the Barthel index, and the PROMIS Applied Cognition-Abilities-Short Form 4a survey. RESULTS: Postoperative delirium was observed in 162 (26%) of 619 patients assessed. When we compared patients with no EEG suppression with those divided into quartiles based on duration of EEG suppression, patients with more suppression were more likely to experience delirium (χ(4) = 25, P < 0.0001). This effect remained significant after we adjusted for potential confounders (odds ratio for log(EEG suppression) 1.22 [99% confidence interval, 1.06-1.40, P = 0.0002] per 1-minute increase in suppression). EEG suppression may have been associated with reduced functional independence (Spearman partial correlation coefficient -0.15, P = 0.02) but not with changes in quality of life or cognitive ability. Predictors of EEG suppression included greater end-tidal volatile anesthetic concentration and lower intraoperative opioid dose. CONCLUSIONS: EEG suppression is an independent risk factor for postoperative delirium. Future studies should investigate whether anesthesia titration to minimize EEG suppression decreases the incidence of postoperative delirium. This is a substudy of the Systematic Assessment and Targeted Improvement of Services Following Yearlong Surgical Outcomes Surveys (SATISFY-SOS) surgical outcomes registry (NCT02032030).