Clinical Decision Model Research Articles

Presentation and physical therapy management using a neuroplasticity approach for patients with hypermobility-related upper cervical instability: a brief report

BackgroundUpper cervical instability (UCI) is a potentially disabling complication of the connective tissue disorders hypermobile Ehlers-Danlos Syndrome and Hypermobility Spectrum Disorders (hEDS/HSD). UCI can impact various neurological structures, including the brainstem, spinal cord, cranial nerves, and blood supply to and from the brain, resulting in complex neurological signs and symptoms in this population. The current study was an observational study applying recent expert consensus recommendations for physical therapy assessment and management of patients with UCI associated with hEDS/HSD.MethodsThis was a retrospective observational study describing how the clinical decision-making model was used to screen, examine, and treat three patients with highly irritable hEDS/HSD-related UCI, resulting in complex neurological presentation. The treatment used a neuroplasticity approach, including proprioception and motor control training emphasizing patient education and biofeedback. Outcome measures tracked progress.ResultsAll patients started with significant disability associated with UCI. One patient returned to full function with intermittent flares that he was able to manage. The second patient continued to have mild-moderate irritability but returned to parenting responsibilities and full-time work. The third patient required cervical fusion and remained disabled but was better able to minimize flares. The number of initial red and yellow flags was associated with the final outcomes, suggesting that the decision-making model might be useful for predicting patient prognosis.ConclusionThis brief report applies recent recommendations for safely evaluating and managing hypermobility-related UCI and provides a first step in experimental studies to test both the assessment and physical therapy treatment approaches.

Optimizing the fairness of survival prediction models for racial/ethnic subgroups: A study on predicting post-operative survival in stage IA and IB non-small cell lung cancer.

380 Background: The recent surge of utilizing machine learning (ML) to develop prediction models for clinical decision-making aids is promising. However, these models can demonstrate racial bias due to inequities in real-world training data. In lung cancer, multiple models have been developed to predict prognosis, but none have been optimized to mitigate bias in performance among racial/ethnic subgroups. We developed a ML model to predict five-year survival in Stage 1A-1B non-small cell lung cancer (NSCLC), ensuring fairness on race. Methods: In the National Cancer Database, we identified patients with histopathologically confirmed stage 1A -1B NSCLC who underwent curative intent lobectomy from 2004 – 2017. We split the study cohort into a training and test sets (70%/30%). We trained and compared various ML models to predict 5-year overall survival. Patient demographic, clinical, and disease characteristics were used as input features for the models. To evaluate model fairness, we used the equalized odds ratio (eOR), which compares the true positive and false positive rates across groups; an eOR value of 1 represents equivalent rates across racial groups. We utilized 3 approaches to mitigate model bias and optimize for fairness of the best “naïve” model: grid search, threshold optimizer, and the exponentiated gradient methods. We evaluated model performance before and after bias mitigation using the area under the curve (AUC). Results: 124,298 patients fit our inclusion/exclusion criteria; 87% of patients were White, 8% were Black/African American, 3% Hispanic, and 2% Asian. Eighty percent of patients were diagnosed with stage 1A cancer; 20% had stage 1B cancer. The best naïve ML model, not optimized for fairness on race, had an eOR of 0.25 with an AUC of 0.66 (95% CI 0.65-0.66) overall. This model demonstrated an AUC of 0.65 (0.65-0.66) among white patients, 0.64 (0.62-0.66) among Black patients, 0.64 (0.60-0.68) among Asian patients, and 0.71 (0.68-0.74) among Hispanic patients. The threshold optimizer bias mitigation strategy improved fairness the most while maintaining similar overall performance of AUC 0.65 (0.64-0.66). With this strategy the eOR improved to 0.83 while AUC remained relatively stable across racial subgroups. Conclusions: We developed a ML model to predict 5-year survival in patients undergoing surgery for stage IA-IB NSCLC and employed model bias mitigation strategies that significantly improved model fairness, without diminishing overall performance. These strategies should be considered when developing prediction models for clinical decision making to avoid perpetuating disparities in care due to algorithm bias. Model performance metrics. Equalized Odds Ratio True Positive Rate False Positive Rate AUC Naive model 0.25 0.61 0.30 0.66 Threshold optimizer mitigated model 0.83 0.62 0.32 0.65

Clinical decision-making during childbirth in health facilities from the perspectives of labouring women, relatives, and health care providers: A scoping review

ProblemFor health care providers to ensure appropriate decision-making in clinical settings during childbirth, facilitators and barriers must be identified. BackgroundWomen who experience a sense of control by participating in the decision-making process, are more likely to have a positive birth experience. However, decision-making may involve hierarchies of close observation and control. AimThe aim of the scoping review was to map and summarise existing literature on the process of clinical decision-making during childbirth from the perspective of labouring women, relatives and health care providers. MethodsWe carried out a scoping review in line with Joanna Briggs Institute scoping review methodology. The search identified studies in Scandinavian or English languages from 2010 - Jan 2023 comprising evidence at different levels of the pyramid, resulting in 18.227 hits. Following the PRISMA checklist, the final inclusion comprised 62 papers. FindingsFour main categories summarized the importance of the following factors: 1) Woman-caregiver relationship, with sub-categories The importance of communication and Midwifery care, 2) Consent and legal issues, 3) Organization, with sub-categories Medicalization, Working atmosphere, and Complexity, and 4) Decision-making tools and models, with sub-categories Shared decision-making, and Other tools and models for decision-making. ConclusionBalancing intuition and expertise of caregivers with evidence-based practices, is crucial to ensure women's participation in decision-making. Furthermore, a trusting relationship between the mother, partner, and health care provider is of utmost importance. Shared decision-making, which appeared to be the primary model for clinical decision-making regardless context, requires reflective practice and is a communication strategy.

Clinician perspectives and recommendations regarding design of clinical prediction models for deteriorating patients in acute care

BackgroundSuccessful deployment of clinical prediction models for clinical deterioration relates not only to predictive performance but to integration into the decision making process. Models may demonstrate good discrimination and calibration, but fail to match the needs of practising acute care clinicians who receive, interpret, and act upon model outputs or alerts. We sought to understand how prediction models for clinical deterioration, also known as early warning scores (EWS), influence the decision-making of clinicians who regularly use them and elicit their perspectives on model design to guide future deterioration model development and implementation.MethodsNurses and doctors who regularly receive or respond to EWS alerts in two digital metropolitan hospitals were interviewed for up to one hour between February 2022 and March 2023 using semi-structured formats. We grouped interview data into sub-themes and then into general themes using reflexive thematic analysis. Themes were then mapped to a model of clinical decision making using deductive framework mapping to develop a set of practical recommendations for future deterioration model development and deployment.ResultsFifteen nurses (n = 8) and doctors (n = 7) were interviewed for a mean duration of 42 min. Participants emphasised the importance of using predictive tools for supporting rather than supplanting critical thinking, avoiding over-protocolising care, incorporating important contextual information and focusing on how clinicians generate, test, and select diagnostic hypotheses when managing deteriorating patients. These themes were incorporated into a conceptual model which informed recommendations that clinical deterioration prediction models demonstrate transparency and interactivity, generate outputs tailored to the tasks and responsibilities of end-users, avoid priming clinicians with potential diagnoses before patients were physically assessed, and support the process of deciding upon subsequent management.ConclusionsPrediction models for deteriorating inpatients may be more impactful if they are designed in accordance with the decision-making processes of acute care clinicians. Models should produce actionable outputs that assist with, rather than supplant, critical thinking.

Risk factors for the development of refeeding syndrome in adults: A systematic review.

Identifying patients with a particularly high risk of refeeding syndrome (RFS) is essential for taking preventive measures. To guide the development of clinical decision-making and risk prediction models or other screening tools for RFS, increased knowledge of risk factors is needed. Therefore, we conducted a systematic review to identify risk factors for the development of RFS. PubMed, EMBASE, Cochrane Library, and Web of Science were searched from January 1990 until March 2023. Studies investigating demographic, clinical, drug use, laboratory, and/or nutrition factors for RFS were considered. The Newcastle-Ottawa Scale was used to appraise the methodological quality of included studies. Of 1589 identified records, 30 studies were included. Thirty-three factors associated with increased risk of RFS after multivariable adjustments were identified. The following factors were reported by two or more studies, with 0-1 study reporting null findings: a previous history of alcohol misuse, cancer, comorbid hypertension, high Acute Physiology and Chronic Health Evaluation II score, high Sequential Organ Failure Assessment score, low Glasgow coma scale score, the use of diuretics before refeeding, low baseline serum prealbumin level, high baseline level of creatinine, and enteral nutrition. The majority of the studies (20, 66.7%) were of high methodological quality. In conclusion, this systematic review informs on several risk factors for RFS in patients. To improve risk stratification and guide development of risk prediction models or other screening tools, further confirmation is needed because there were a small number of studies and a low number of high-quality studies on each factor.

A predictive machine-learning model for clinical decision-making in washed microbiota transplantation on ulcerative colitis

Background and AimMachine learning based on clinical data and treatment protocols for better clinical decision-making is a current research hotspot. This study aimed to build a machine learning model on washed microbiota transplantation (WMT) for ulcerative colitis (UC), providing patients and clinicians with a new evaluation system to optimize clinical decision-making.MethodsPatients with UC who underwent WMT via mid-gut or colonic delivery route at an affiliated hospital of Nanjing Medical University from April 2013 to June 2022 were recruited. Model ensembles based on the clinical indicators were constructed by machine-learning to predict the clinical response of WMT after one month.ResultsA total of 366 patients were enrolled in this study, with 210 patients allocated for training and internal validation, and 156 patients for external validation. The low level of indirect bilirubin, activated antithrombin III, defecation frequency and cholinesterase and the elderly and high level of creatine kinase, HCO3- and thrombin time were related to the clinical response of WMT at one month. Besides, the voting ensembles exhibited an area under curve (AUC) of 0.769 ± 0.019 [accuracy, 0.754; F1-score, 0.845] in the internal validation; the AUC of the external validation was 0.614 ± 0.017 [accuracy, 0.801; F1-score, 0.887]. Additionally, the model was available at https://wmtpredict.streamlit.app.ConclusionsThis study pioneered the development of a machine learning model to predict the one-month clinical response of WMT on UC. The findings demonstrate the potential value of machine learning applications in the field of WMT, opening new avenues for personalized treatment strategies in gastrointestinal disorders.Trial registrationclinical trials, NCT01790061. Registered 09 February 2013 - Retrospectively registered, https://clinicaltrials.gov/study/NCT01790061

Consideration for ICU Admission in New-onset Encephalitis Patients: A Clinical Decision Model Development and Validation Study Based on Two Multicenter Cohorts (P9-2.002)

Consideration for ICU Admission in New-onset Encephalitis Patients: A Clinical Decision Model Development and Validation Study Based on Two Multicenter Cohorts (P9-2.002)

Perspectives of Oncologists on the Ethical Implications of Using Artificial Intelligence for Cancer Care

Artificial intelligence (AI) tools are rapidly integrating into cancer care. Understanding stakeholder views on ethical issues associated with the implementation of AI in oncology is critical to optimal deployment. To evaluate oncologists' views on the ethical domains of the use of AI in clinical care, including familiarity, predictions, explainability (the ability to explain how a result was determined), bias, deference, and responsibilities. This cross-sectional, population-based survey study was conducted from November 15, 2022, to July 31, 2023, among 204 US-based oncologists identified using the National Plan & Provider Enumeration System. The primary outcome was response to a question asking whether participants agreed or disagreed that patients need to provide informed consent for AI model use during cancer treatment decisions. Of 387 surveys, 204 were completed (response rate, 52.7%). Participants represented 37 states, 120 (63.7%) identified as male, 128 (62.7%) as non-Hispanic White, and 60 (29.4%) were from academic practices; 95 (46.6%) had received some education on AI use in health care, and 45.3% (92 of 203) reported familiarity with clinical decision models. Most participants (84.8% [173 of 204]) reported that AI-based clinical decision models needed to be explainable by oncologists to be used in the clinic; 23.0% (47 of 204) stated they also needed to be explainable by patients. Patient consent for AI model use during treatment decisions was supported by 81.4% of participants (166 of 204). When presented with a scenario in which an AI decision model selected a different treatment regimen than the oncologist planned to recommend, the most common response was to present both options and let the patient decide (36.8% [75 of 204]); respondents from academic settings were more likely than those from other settings to let the patient decide (OR, 2.56; 95% CI, 1.19-5.51). Most respondents (90.7% [185 of 204]) reported that AI developers were responsible for the medico-legal problems associated with AI use. Some agreed that this responsibility was shared by physicians (47.1% [96 of 204]) or hospitals (43.1% [88 of 204]). Finally, most respondents (76.5% [156 of 204]) agreed that oncologists should protect patients from biased AI tools, but only 27.9% (57 of 204) were confident in their ability to identify poorly representative AI models. In this cross-sectional survey study, few oncologists reported that patients needed to understand AI models, but most agreed that patients should consent to their use, and many tasked patients with choosing between physician- and AI-recommended treatment regimens. These findings suggest that the implementation of AI in oncology must include rigorous assessments of its effect on care decisions as well as decisional responsibility when problems related to AI use arise.

Paediatric meningitis in the conjugate vaccine era and a novel clinical decision model to predict bacterial aetiology

The aims of this study were to assess aetiology and clinical characteristics in childhood meningitis, and develop clinical decision rules to distinguish bacterial meningitis from other similar clinical syndromes. Children aged <16 years hospitalised with suspected meningitis/encephalitis were included, and prospectively recruited at 31 UK hospitals. Meningitis was defined as identification of bacteria/viruses from cerebrospinal fluid (CSF) and/or a raised CSF white blood cell count. New clinical decision rules were developed to distinguish bacterial from viral meningitis and those of alternative aetiology. The cohort included 3002 children (median age 2·4 months); 1101/3002 (36·7%) had meningitis, including 180 bacterial, 423 viral and 280 with no pathogen identified. Enterovirus was the most common pathogen in those aged <6 months and 10-16 years, with Neisseria meningitidis and/or Streptococcus pneumoniae commonest at age 6 months to 9 years. The Bacterial Meningitis Score had a negative predictive value of 95·3%. We developed two clinical decision rules, that could be used either before (sensitivity 82%, specificity 71%) or after lumbar puncture (sensitivity 84%, specificity 93%), to determine risk of bacterial meningitis. Bacterial meningitis comprised 6% of children with suspected meningitis/encephalitis. Our clinical decision rules provide potential novel approaches to assist with identifying children with bacterial meningitis. This study was funded by the Meningitis Research Foundation, Pfizer and the NIHR Programme Grants for Applied Research.

Evaluating the clinical utility of an easily applicable prediction model of suicide attempts, newly developed and validated with a general community sample of adults.

A suicide attempt (SA) is a clinically serious action. Researchers have argued that reducing long-term SA risk may be possible, provided that at-risk individuals are identified and receive adequate treatment. Algorithms may accurately identify at-risk individuals. However, the clinical utility of algorithmically estimated long-term SA risk has never been the predominant focus of any study. The data of this report stem from CoLaus|PsyCoLaus, a prospective longitudinal study of general community adults from Lausanne, Switzerland. Participants (N = 4,097; Mage = 54 years, range: 36-86; 54% female) were assessed up to four times, starting in 2003, approximately every 4-5 years. Long-term individual SA risk was prospectively predicted, using logistic regression. This algorithm's clinical utility was assessed by net benefit (NB). Clinical utility expresses a tool's benefit after having taken this tool's potential harm into account. Net benefit is obtained, first, by weighing the false positives, e.g., 400 individuals, at the risk threshold, e.g., 1%, using its odds (odds of 1% yields 1/(100-1) = 1/99), then by subtracting the result (400*1/99 = 4.04) from the true positives, e.g., 5 individuals (5-4.04), and by dividing the result (0.96) by the sample size, e.g., 800 (0.96/800). All results are based on 100 internal cross-validations. The predictors used in this study were: lifetime SA, any lifetime mental disorder, sex, and age. SA at any of the three follow-up study assessments was reported by 1.2%. For a range of seven a priori selected threshold probabilities, ranging between 0.5% and 2%, logistic regression showed highest overall NB in 97.4% of all 700 internal cross-validations (100 for each selected threshold probability). Despite the strong class imbalance of the outcome (98.8% no, 1.2% yes) and only four predictors, clinical utility was observed. That is, using the logistic regression model for clinical decision making provided the most true positives, without an increase of false positives, compared to all competing decision strategies. Clinical utility is one among several important prerequisites of implementing an algorithm in routine practice, and may possibly guide a clinicians' treatment decision making to reduce long-term individual SA risk. The novel metric NB may become a standard performance measure, because the a priori invested clinical considerations enable clinicians to interpret the results directly.

Differential genetic mutations and immune cell infiltration in high- and low-risk STAD: Implications for prognosis and immunotherapy efficacy.

This study investigates genetic mutations and immune cell dynamics in stomach adenocarcinoma (STAD), focusing on identifying prognostic markers and therapeutic targets. Analysis of TCGA-STAD samples revealed C > A as the most common single nucleotide variant (SNV) in both high and low-risk groups. Key mutated driver genes included TTN, TP53 and MUC16, with frame-shift mutations more prevalent in the low-risk group and missense mutations in the high-risk group. Interaction analysis of hub genes such as C1QA and CD68 showed significant correlations, impacting immune cell infiltration patterns. Using ssGSEA, we found higher immune cell infiltration (B cells, CD4+ T cells, CD8+ T cells, DC cells, NK cells) in the high-risk group, correlated with increased risk scores. xCell algorithm results indicated distinct immune infiltration levels between the groups. The study's risk scoring model proved effective in prognosis prediction and immunotherapy efficacy assessment. Key molecules like CD28, CD27 and SLAMF7 correlated significantly with risk scores, suggesting potential targets for high-risk STAD patients. Drug sensitivity analysis showed a negative correlation between risk scores and sensitivity to certain treatments, indicating potential therapeutic options for high-risk STAD patients. We also validated the carcinogenic role of RPL14 in gastric cancer through phenotypic experiments, demonstrating its influence on cancer cell proliferation, invasion and migration. Overall, this research provides crucial insights into the genetic and immune aspects of STAD, highlighting the importance of a risk scoring model for personalized treatment strategies and clinical decision-making in gastric cancer management.

Decomposed FDG PET-based phenotypic heterogeneity predicting clinical prognosis and decision-making in temporal lobe epilepsy patients.

This study utilized a data-driven Bayesian model to automatically identify distinct latent disease factors represented by overlapping glucose metabolism patterns from 18F-Fluorodeoxyglucose PET (18F-FDG PET) to analyze heterogeneity among patients with TLE. We employed unsupervised machine learning to estimate latent disease factors from 18F-FDG PET scans, representing whole-brain glucose metabolism patterns in seventy patients with TLE. We estimated the extent to which multiple distinct factors were expressed within each participant and analyzed their relevance to epilepsy burden, including seizure onset, duration, and frequency. Additionally, we established a predictive model for clinical prognosis and decision-making. We identified three latent disease factors: hypometabolism in the unilateral temporal lobe and hippocampus (factor 1), hypometabolism in bilateral prefrontal lobes (factor 2), and hypometabolism in bilateral temporal lobes (factor 3), variably co-expressed within each patient. Factor 3 demonstrated the strongest negative correlation with the age of onset and duration (r = - 0.33, - 0.38 respectively, P < 0.05). The supervised classifier, trained on latent disease factors for predicting patient-specific antiepileptic drug (AED) responses, achieved an area under the curve (AUC) of 0.655. For post-surgical seizure outcomes, the AUC was 0.857, and for clinical decision-making, it was 0.965. Decomposing 18F-FDG PET-based phenotypic heterogeneity facilitates individual-level predictions relevant to disease monitoring and personalized therapeutic strategies.

Risk subgroups and intervention effects among infants at high risk for peanut allergy: A model for clinical decision making.

The Learning Early About Peanut Allergy (LEAP) trial showed that early dietary introduction of peanut reduced the risk of developing peanut allergy by age 60 months in infants at high risk for peanut allergy. In this secondary analysis of LEAP data, we aimed to determine risk subgroups within these infants and estimate their respective intervention effects of early peanut introduction. LEAP raw data were retrieved from ITNTrialShare.org. Conditional random forest was applied to participants in the peanut avoidance arm to select statistically important features for the classification and regression tree (CART) analysis to group infants based on their risk of peanut allergy at 60 months of age. Intervention effects were estimated for each derived risk subgroup using data from both arms. Our main model was generated based on baseline data when the participants were 4-11 months old. Specific IgE measurements were truncated to account for the limit of detection commonly used by laboratories in clinical practice. The model found infants with higher predicted probability of peanut allergy at 60 months of age had a similar relative risk reduction, but a greater absolute risk reduction in peanut allergy with early introduction of peanut, than those with lower probability. The intervention effects were significant across all risk subgroups. Participants with baseline peanut sIgE ≥0.22 kU/L (n = 78) had an absolute risk reduction of 40.4% (95% CI 27.3, 51.9) whereas participants with baseline peanut sIgE<0.22 kU/L and baseline Ara h 2 sIgE <0.10 kU/L (n = 226) had an absolute risk reduction of 6.5% (95% CI 2.6, 11.0). These findings were consistent in sensitivity analyses using alternative models. In this study, risk subgroups were determined among infants from the LEAP trial based on the probability of developing peanut allergy and the intervention effects of early peanut introduction were estimated. This may be relevant for further risk assessment and personalized clinical decision-making.

Prognostic Significance of Glycolysis-Related Genes in Lung Squamous Cell Carcinoma.

Lung squamous cell carcinoma (LUSC) is one of the most common malignancies. There is growing evidence that glycolysis-related genes play a critical role in tumor development, maintenance, and therapeutic response by altering tumor metabolism and thereby influencing the tumor immune microenvironment. However, the overall impact of glycolysis-related genes on the prognostic significance, tumor microenvironment characteristics, and treatment outcome of patients with LUSC has not been fully elucidated. We used The Cancer Genome Atlas (TCGA) dataset to screen glycolysis-related genes with prognostic effects in LUSC and constructed signature and nomogram models using Lasso and Cox regression, respectively. In addition, we analyzed the immune infiltration and tumor mutation load of the genes in the models. We finally obtained a total of glycolysis-associated DEGs. The signature model and nomogram model had good prognostic power for LUSC. Gene expression in the models was highly correlated with multiple immune cells in LUSC. Through this analysis, we have identified and validated for the first time that glycolysis-related genes are highly associated with the development of LUSC. In addition, we constructed the signature model and nomogram model for clinical decision-making.

Nurses’ clinical decision-making models in the care of older adults: A cross-sectional study

Clinical decision-making is referred to as a correct judgment based on knowledge and cognitive skills. It has an effective role in providing safe and effective nursing care. Often these decision-making skills are based on critical thinking and evidence-based practice. Therefore, due to the importance of the present issue, this study aimed to assess the nurses’ clinical decision-making models in the care of older adults. This cross-sectional study was conducted on 384 nurses working in the teaching hospitals of Gorgan University of Medical Sciences in 2021. Data was collected by the standard questionnaire of clinical decision-making model. Pearson's correlation coefficient test was used to evaluate the relationship between clinical decision making and study variables.

A National Survey of Oncologists' Perspectives on the Ethical Implications of Artificial Intelligence in Cancer Care

Introduction Artificial Intelligence (AI) is an emerging technology that promises advancements across blood cancer discovery and care delivery. There are already FDA-approved uses of AI in hematology and oncology but few data available about its use or utility. Concerns have been raised over AI's potential biases, “explainability,” and who is responsible for its use. Oncologists' familiarity with AI and perspectives on these issues are not well known and are critical to understanding the ethical deployment of AI in blood cancer care. Methods From January to July 2023, we performed a national cross-sectional survey of practicing oncologists. A draft survey instrument was developed by a team with content and methodological expertise. Iterative refinement occurred through cognitive debriefing with 5 practicing hematologist/oncologists. The final instrument included 24 questions covering domains of AI familiarity, predictions, explainability, and responsibilities. A random sample of US oncologists was identified using the National Plan &amp; Provider Enumeration System. Paper surveys were mailed with $25 incentivizing gift cards. Reminder letters and phone calls were used for non-responders with an electronic survey option provided. Responses were analyzed descriptively. X2 testing assessed bivariate associations between responses and participant practice setting, with odds ratios (OR) reported. Results Of 399 mailed surveys, 12 were undeliverable and 201 were completed for a response rate of 51.9%. Question missingness was &amp;lt;1%. Participants were 63.2% White, 16.9% Indian Asian, 10.0% Eastern Asian or Pacific Islander, and 4.5% Black; 6.0% were Hispanic. They represented 37 states. The majority were male (63.7%), 40-59 years old (54.2%), and had no specific training in AI (52.2%); 68.2% had at least 11 years in practice and 29.9% practiced at a primary academic hospital. Questions assessing overall familiarity with and predictions related to AI are shown in Table 1. Those at primary academic hospitals were more likely to be familiar with radiology models (OR 2.82, p&amp;lt;0.001) but not pathology, prognostic, or clinical decision models (all p&amp;gt;0.05). They were more likely to predict that AI would improve side effect management (OR 2.01, p=0.03) and end-of-life decision making (OR 2.07, p=0.02). High proportions of respondents thought that AI prognostic (67.1%) and clinical decision models (62.1%) needed to be explainable by clinicians and researchers to be used in clinic. When posed with a scenario where an FDA-approved AI model selects a different regimen than the oncologist planned to propose (Figure 1), the largest proportion reported they would present both options and let the patient decide (37.3%). Of those recommending a regimen, respondents from primary academic hospitals more frequently recommend the AI-generated regimen (56.1%), while respondents from elsewhere recommended their original regimen (69.0%; OR 2.85, p=0.007). Acceptability of some direct-to-patient AI model applications is shown in Table 1. A majority of respondents (90.5%) thought that AI developers should be responsible for medico-legal problems caused by AI use; 47.8% thought that physicians using the tool should be responsible. Though most (76.1%) agreed that it was oncologists' responsibility to protect patients from biased AI tools, few (27.3%) were confident in their ability to identify how representative the data used in an AI model was relative to the population with that cancer (including 66.7% of those who agreed it was oncologists' responsibility). More respondents thought that patients should need to consent to the use of AI tools in cancer treatment decisions (81.1%) than cancer diagnosis (55.7%); those from primary academic hospitals were less likely to respond that patients should consent for treatment or diagnosis (OR 0.43, p=0.02 and OR 0.44, p=0.009). Conclusions In this national survey of oncologists' views on AI, there was optimism about its impact on care with notable caveats related to end-of-life care, patient trust, and care disparities. Despite many approving use of AI models too complex for patients to explain, a substantial proportion deferred to patient choice when AI and physician recommendations differed. There was also dissonance between accepting responsibility to protect patients from bias and respondents' self-perceived ability to do so.

Federated Inverse Reinforcement Learning for Smart ICUs With Differential Privacy

Clinical decision-making models have been developed to support therapeutic interventions based on medical data from either a single hospital or multiple hospitals. However, models based on multi-hospital data require collaboration among hospitals to integrate local data, which can result in information leakage and violate patient privacy. To address this challenge, we propose a novel approach that combines federated learning with inverse reinforcement learning to create an efficient medical decision-making support tool while preserving patient privacy. Our approach uses an inverse reinforcement learning algorithm with differential privacy to train a neural network-based agent on local data containing clinician trajectories, which learns a private treatment policy by observing patients’ conditions. Additionally, we integrate federated learning into the proposed algorithm to learn a global optimal action policy collaboratively among various smart ICUs, overcoming data limitations at each hospital. We evaluate our approach using real-world medical data and demonstrate that it achieves superior performance in a distributed manner.

Cognitive processes in pharmacists’ clinical decision-making

BackgroundPharmacists’ clinical decision-making is a core process in pharmaceutical care. However, the practical aspects and effective teaching methods of this process remain largely unexplored. ObjectiveTo examine the cognitive processes involved in pharmacists’ perceptions of how they make clinical decisions in pharmacy practice. MethodsSemi-structured, face-to-face interviews were conducted with pharmacists working in community, outpatient, and hospital care in the Netherlands between August and December 2021. Participants were explicitly asked for examples when asked how they make clinical decisions in practice and how they teach this to others. After transcribing audio-recorded interviews, an inductive thematic analysis was conducted to identify cognitive processes. A theoretical model of clinical decision-making was then used and adapted to structure the identified processes. ResultsIn total, 21 cognitive processes were identified from interviews with 16 pharmacists working in community (n = 5), outpatient (n = 2), and hospital care (n = 9). These cognitive processes were organized into 8 steps of the adapted theoretical model, i.e. problem and demand for care consideration, information collection, clinical reasoning, clinical judgment, shared decision-making, implementation, outcomes evaluation, and reflection. Pharmacists struggled to articulate their clinical decision-making and went back-and-forth in their explanations of this process. All pharmacists emphasized the importance of identifying the problem and described how they collect information through reviewing, gathering, recalling, and investigating. Clinical reasoning entailed various cognitive processes, of which comprehending the problem in the patient's context was deemed challenging at times. Pharmacists seemed least active in evaluating patient outcomes and reflecting on these outcomes. ConclusionsPharmacists use multiple cognitive processes when making clinical decisions in pharmacy practice, and their back-and-forth explanations emphasize its dynamic nature. This study adds to a greater understanding of how pharmacists make clinical decisions and to the development of a theoretical model that describes this process, which can be used in pharmacy practice and education.

Risk Stratification of Postoperative Adjuvant Therapy for Endometrial Cancer (POAT-ENDORISK) Based on Bayesian Network Model: A Development and Validation Study

To establish a Bayesian network (BN) model for postoperative adjuvant treatment of early endometrial carcinoma (EC) patients. We retrospectively analyzed the data of 1280 early EC patients treated by multiple institutions in China from 1999 to 2017. All patients received primary hysterectomy/bilateral salpingo-oophorectomy and adjuvant radiotherapy. The FIGO 2009 stage of all patients is stage I and stage II EC, and the median age is 57 years old. All patients are grouped according to the ESMO-ESGO-ESTRO risk stratification. The clinicopathologic factors, treatment-related factors, local regional recurrence, distant metastasis and cancer-specific survival rate (CSS) of all patients were reviewed. We divide the original data set into training set and Validation set according to the ratio of 7:3. The training of the Bayesian network model is completed on Netica, and the test of the model effect is finally completed on the test set. After variable screening, a total of 14 characteristic variables entered the final model. A total of 896 patients were used for the development of BN model, and 384 patients were used for the validation of BN model. The results of the model showed that the factors directly related to CSS were locoregional failure (LRF), radiotherapy mode, distant metastasis (DM). Factors directly related to DM were chemotherapy, LRF, CSS. The factors directly related to LRF were risk stratification, preoperative serum CA125 and preoperative HB. The accuracy, sensibility, specificity, micro-f1, micro-f1, weighted-f1 and AUC of BN model in predicting DM and CSS were better than XGBoost model. In this study, we integrated almost all clinical pathology and treatment information related to postoperative adjuvant treatment of early EC patients and established a BN model for personalized clinical decision-making of postoperative adjuvant treatment of early EC patients. The results showed the complex correlation among the variables, and the overall prediction ability and visualization effect of BN model was significantly better than XGBoost model. Prospective research is needed before clinical implementation.

A surgical Decision-making scoring model for spontaneous ventilation- and mechanical ventilation-video-assisted thoracoscopic surgery in non-small-cell lung cancer patients

BackgroundsSpontaneous ventilation-video-assisted thoracoscopic surgery (SV-VATS) has been applied to non-small cell lung cancer (NSCLC) patients in many centers. Since it remains a new and challenging surgical technique, only selected patients can be performed SV-VATS. We aim to conduct a retrospective single-center study to develop a clinical decision-making model to make surgery decision between SV-VATS and MV (mechanical ventilation) -VATS in NSCLC patients more objectively and individually.MethodsFour thousand three hundred sixty-eight NSCLC patients undergoing SV-VATS or MV-VATS in the department of thoracic surgery between 2011 and 2018 were included. Univariate and multivariate regression analysis were used to identify potential factors influencing the surgical decisions. Factors with statistical significance were selected for constructing the Surgical Decision-making Scoring (SDS) model. The performance of the model was validated by area under the receiver operating characteristic curve (AUC), calibration curves and decision curve analysis (DCA).ResultsThe Surgical Decision-making Scoring (SDS) model was built guided by the clinical judgment and statistically significant results of univariate and multivariate regression analyses of potential predictors, including smoking status (p = 0.03), BMI (p < 0.001), ACCI (p = 0.04), T stage (p < 0.001), N stage (p < 0.001), ASA grade (p < 0.001) and surgical technique (p < 0.001). The AUC of the training group and the testing group were 0.72 and 0.70, respectively. The calibration curves and the DCA curve revealed that the SDS model has a desired performance in predicting the surgical decision.ConclusionsThis SDS model is the first clinical decision-making model developed for an individual NSCLC patient to make decision between SV-VATS and MV-VATS.