BMC Medical Informatics and Decision Making最新文献

Background: To evaluate the accuracy, reliability, quality, and readability of responses generated by ChatGPT-3.5, ChatGPT-4, Gemini, and Copilot in relation to orthodontic clear aligners.

Methods: Frequently asked questions by patients/laypersons about clear aligners on websites were identified using the Google search tool and these questions were posed to ChatGPT-3.5, ChatGPT-4, Gemini, and Copilot AI models. Responses were assessed using a five-point Likert scale for accuracy, the modified DISCERN scale for reliability, the Global Quality Scale (GQS) for quality, and the Flesch Reading Ease Score (FRES) for readability.

Results: ChatGPT-4 responses had the highest mean Likert score (4.5 ± 0.61), followed by Copilot (4.35 ± 0.81), ChatGPT-3.5 (4.15 ± 0.75) and Gemini (4.1 ± 0.72). The difference between the Likert scores of the chatbot models was not statistically significant (p > 0.05). Copilot had a significantly higher modified DISCERN and GQS score compared to both Gemini, ChatGPT-4 and ChatGPT-3.5 (p < 0.05). Gemini's modified DISCERN and GQS score was statistically higher than ChatGPT-3.5 (p < 0.05). Gemini also had a significantly higher FRES compared to both ChatGPT-4, Copilot and ChatGPT-3.5 (p < 0.05). The mean FRES was 38.39 ± 11.56 for ChatGPT-3.5, 43.88 ± 10.13 for ChatGPT-4 and 41.72 ± 10.74 for Copilot, indicating that the responses were difficult to read according to the reading level. The mean FRES for Gemini is 54.12 ± 10.27, indicating that Gemini's responses are more readable than other chatbots.

Conclusions: All chatbot models provided generally accurate, moderate reliable and moderate to good quality answers to questions about the clear aligners. Furthermore, the readability of the responses was difficult. ChatGPT, Gemini and Copilot have significant potential as patient information tools in orthodontics, however, to be fully effective they need to be supplemented with more evidence-based information and improved readability.

Background: Sub-Saharan Africa bears the highest burden of sickle cell disease (SCD) globally with Nigeria, Democratic Republic of Congo, Tanzania, Uganda being the most affected countries. Uganda reports approximately 20,000 SCD births annually, constituting 6.67% of reported global SCD births. Despite this, there is a paucity of comprehensive data on SCD from the African continent. SCD registries offer a promising avenue for conducting prospective studies, elucidating disease severity patterns, and evaluating the intricate interplay of social, environmental, and genetic factors. This paper describes the establishment of the Sickle Pan Africa Research Consortium (SPARCo) Uganda registry, encompassing its design, development, data collection, and key insights learned, aligning with collaborative efforts in Nigeria, Tanzania, and Ghana SPARCo registries.

Methods: The registry was created using pre-existing case report forms harmonized from the SPARCo data dictionary and ontology to fit Uganda clinical needs. The case report forms were developed with SCD data elements of interest including demographics, consent, baseline, clinical, laboratory and others. That data was then parsed into a customized REDCap database, configured to suit the optimized ontologies and support retrieval aggregations and analyses. Patients were enrolled from one national referral and three regional referral hospitals in Uganda.

Results: A nationwide electronic patient-consented registry for SCD was established from four regional hospitals. A total of 5,655 patients were enrolled from Mulago National Referral Hospital (58%), Jinja Regional Referral (14.4%), Mbale Regional Referral (16.9%), and Lira Regional Referral (10.7%) hospitals between June 2022 and October 2023.

Conclusion: Uganda has been able to develop a SCD registry consistent with data from Tanzania, Nigeria and Ghana. Our findings demonstrate that it's feasible to develop longitudinal SCD registries in sub-Saharan Africa. These registries will be crucial for facilitating a range of studies, including the analysis of SCD clinical phenotypes and patient outcomes, newborn screening, and evaluation of hydroxyurea use, among others. This initiative underscores the potential for developing comprehensive disease registries in resource-limited settings, fostering collaborative, data-driven research efforts aimed at addressing the multifaceted challenges of SCD in Africa.

Background: This study aims to predict the trend of procurement and storage of various blood products, as well as planning and monitoring the consumption of blood products in different centers across Iran based on artificial intelligence until the year 2027.

Methods: This research constitutes a time-series investigation within the realm of longitudinal studies. In this study, information on the number of packed red blood cells (RBC), leukoreduced red blood cells (LR-RBC), and platelets (PLT), PLT-Apheresis, and fresh frozen plasma (FFP) was requested from all blood transfusion centers in the country and extracted using a unified protocol. After the initial examination of the information and addressing data issues and inconsistencies, the corrected data were analyzed. Both conventional and artificial intelligence approaches were used to predict each product in this study. The best model was selected based on goodness-of-fit indicators RMSE and MAPE.

Results: Based on the obtained results, the FFP product will follow a relatively consistent process similar to previous years in the next five years. The PLT product is predicted to have a growing trend over the next 5 years, which applies to both the demand and supply of the product. The PLT-Apheresis product also shows a similar upward trend, albeit with a lower growth rate. The RBC product will have a constant trend over a 5-year period (long-term) according to both models, taking into account short-term changes. Similarly, there is a similar trend in LR-RBC, with the expectation that short-term pattern repetition will continue over a 5-year period (long-term). Comparing the goodness-of-fit results, the LSTM model proved to be better for predicting the dominant blood products.

Conclusions: The growth of the elderly population and diseases related to old age, and on the other hand, the trend of increasing the consumption of the product with a short lifespan (PLT) requires the activation of the management of the patient's blood, especially in relation to this product in medical centers. The trend for other products in the next five years is similar to previous years, and no growth in demand is observed. The LSTM method, considering periodic and cyclical events, has performed the prediction.

Background: The National Institute of Health and Social Care Research (NIHR) Health Informatics Collaborative (HIC) for Hearing Health has been established in the UK to curate routinely collected hearing health data to address research questions. This study defines priority research areas, outlines its aims, governance structure and demonstrates how hearing health data have been integrated into a common data model using pure tone audiometry (PTA) as a case study.

Methods: After identifying key research aims in hearing health, the governance structure for the NIHR HIC for Hearing Health is described. The Observational Medical Outcomes Partnership (OMOP) was chosen as our common data model to provide a case study example.

Results: The NIHR HIC Hearing Health theme have developed a data architecture outlying the flow of data from all of the various siloed electronic patient record systems to allow the effective linkage of data from electronic patient record systems to research systems. Using PTAs as an example, OMOPification of hearing health data successfully collated a rich breadth of datapoints across multiple centres.

Conclusion: This study identified priority research areas where routinely collected hearing health data could be useful. It demonstrates integration and standardisation of such data into a common data model from multiple centres. By describing the process of data sharing across the HIC, we hope to invite more centres to contribute and utilise data to address research questions in hearing health. This national initiative has the power to transform UK hearing research and hearing care using routinely collected clinical data.

Background: A central goal of modern evidence-based medicine is the development of simple and easy to use tools that help clinicians integrate quantitative information into medical decision-making. The Bayesian Pre-test/Post-test Probability (BPP) framework is arguably the most well known of such tools and provides a formal approach to quantify diagnostic uncertainty given the result of a medical test or the presence of a clinical sign. Yet, clinical decision-making goes beyond quantifying diagnostic uncertainty and requires that that uncertainty be balanced against the various costs and benefits associated with each possible decision. Despite increasing attention in recent years, simple and flexible approaches to quantitative clinical decision-making have remained elusive.

Methods: We extend the BPP framework using concepts of Bayesian Decision Theory. By integrating cost, we can expand the BPP framework to allow for clinical decision-making.

Results: We develop a simple quantitative framework for binary clinical decisions (e.g., action/inaction, treat/no-treat, test/no-test). Let p be the pre-test or post-test probability that a patient has disease. We show that $r_{}^{\ast }=(1-p)/p$ represents a critical value called a decision boundary. In terms of the relative cost of under- to over-acting, $r_{}^{\ast }$ represents the critical value at which action and inaction are equally optimal. We demonstrate how this decision boundary can be used at the bedside through case studies and as a research tool through a reanalysis of a recent study which found widespread misestimation of pre-test and post-test probabilities among clinicians.

Conclusions: Our approach is so simple that it should be thought of as a core, yet previously overlooked, part of the BPP framework. Unlike prior approaches to quantitative clinical decision-making, our approach requires little more than a hand-held calculator, is applicable in almost any setting where the BPP framework can be used, and excels in situations where the costs and benefits associated with a particular decision are patient-specific and difficult to quantify.

Deep neural networks (DNN) have fundamentally revolutionized the artificial intelligence (AI) field. The transformer model is a type of DNN that was originally used for the natural language processing tasks and has since gained more and more attention for processing various kinds of sequential data, including biological sequences and structured electronic health records. Along with this development, transformer-based models such as BioBERT, MedBERT, and MassGenie have been trained and deployed by researchers to answer various scientific questions originating in the biomedical domain. In this paper, we review the development and application of transformer models for analyzing various biomedical-related datasets such as biomedical textual data, protein sequences, medical structured-longitudinal data, and biomedical images as well as graphs. Also, we look at explainable AI strategies that help to comprehend the predictions of transformer-based models. Finally, we discuss the limitations and challenges of current models, and point out emerging novel research directions.

Introduction: Sepsis-associated acute kidney injury (SA-AKI) is strongly associated with poor prognosis. We aimed to build a machine learning (ML)-based clinical model to predict 1-year mortality in patients with SA-AKI.

Methods: Six ML algorithms were included to perform model fitting. Feature selection was based on the feature importance evaluated by the SHapley Additive exPlanations (SHAP) values. Area under the receiver operating characteristic curve (AUROC) was used to evaluate the discriminatory ability of the prediction model. Calibration curve and Brier score were employed to assess the calibrated ability. Our ML-based prediction models were validated both internally and externally.

Results: A total of 12,750 patients with SA-AKI and 55 features were included to build the prediction models. We identified the top 10 predictors including age, ICU stay and GCS score based on the feature importance. Among the six ML algorithms, the CatBoost showed the best prediction performance with an AUROC of 0.813 and Brier score of 0.119. In the external validation set, the predictive value remained favorable (AUROC = 0.784).

Conclusion: In this study, we developed and validated a ML-based prediction model based on 10 commonly used clinical features which could accurately and early identify the individuals at high-risk of long-term mortality in patients with SA-AKI.

Objective: Based on the Omaha problem classification system, a sensitivity outcome index system for home nursing of elderly liver transplant patients was established.

Methods: Through a comprehensive literature review and rigorous application of the Delphi method, a panel of 20 experts completed two rounds of effective letter consultation to obtain expert consensus opinions. The contents of indicators were determined based on this process, and the analytic hierarchy process was employed to confirm the weightage assigned to each indicator. Consequently, we established a sensitivity outcome index system for home care in elderly liver transplant patients.

Results: The effective recovery rate of the questionnaire in two rounds of expert consultation was 100%, and the proportion of experts who gave opinions was 55% and 15%, respectively, indicating that the experts were highly active. The expert authority coefficients were calculated as 0.904 and 0.905, respectively, indicating a high degree of expert authority. In the second round, Kendall's coordination coefficients for primary, secondary, and tertiary indicators were determined to be 0.419, 0.418, and 0.394 (P < 0.001), indicating that expert opinions tended to be consistent. Finally, we established a comprehensive sensitivity outcome index system comprising 4 first-level indexes, 20 s-level indexes, and 72 third-level indexes specifically designed for elderly liver transplantation patients.

Conclusion: The sensitivity outcome index system of home nursing for elderly liver transplant patients can provide theoretical basis for nursing staff to build accurate individualized continuous nursing model.

Background: Electronic Health Records (EHR) are widely used to develop clinical prediction models (CPMs). However, one of the challenges is that there is often a degree of informative missing data. For example, laboratory measures are typically taken when a clinician is concerned that there is a need. When data are the so-called Not Missing at Random (NMAR), analytic strategies based on other missingness mechanisms are inappropriate. In this work, we seek to compare the impact of different strategies for handling missing data on CPMs performance.

Methods: We considered a predictive model for rapid inpatient deterioration as an exemplar implementation. This model incorporated twelve laboratory measures with varying levels of missingness. Five labs had missingness rate levels around 50%, and the other seven had missingness levels around 90%. We included them based on the belief that their missingness status can be highly informational for the prediction. In our study, we explicitly compared the various missing data strategies: mean imputation, normal-value imputation, conditional imputation, categorical encoding, and missingness embeddings. Some of these were also combined with the last observation carried forward (LOCF). We implemented logistic LASSO regression, multilayer perceptron (MLP), and long short-term memory (LSTM) models as the downstream classifiers. We compared the AUROC of testing data and used bootstrapping to construct 95% confidence intervals.

Results: We had 105,198 inpatient encounters, with 4.7% having experienced the deterioration outcome of interest. LSTM models generally outperformed other cross-sectional models, where embedding approaches and categorical encoding yielded the best results. For the cross-sectional models, normal-value imputation with LOCF generated the best results.

Conclusion: Strategies that accounted for the possibility of NMAR missing data yielded better model performance than those did not. The embedding method had an advantage as it did not require prior clinical knowledge. Using LOCF could enhance the performance of cross-sectional models but have countereffects in LSTM models.

Despite the high creation cost, annotated corpora are indispensable for robust natural language processing systems. In the clinical field, in addition to annotating medical entities, corpus creators must also remove personally identifiable information (PII). This has become increasingly important in the era of large language models where unwanted memorization can occur. This paper presents a corpus annotated to anonymize personally identifiable information in 1,787 anamneses of work-related accidents and diseases in Spanish. Additionally, we applied a previously released model for Named Entity Recognition (NER) trained on referrals from primary care physicians to identify diseases, body parts, and medications in this work-related text. We analyzed the differences between the models and the gold standard curated by a physician in detail. Moreover, we compared the performance of the NER model on the original narratives, in narratives where personal information has been masked, and in texts where the personal data is replaced by another similar surrogate value (pseudonymization). Within this publication, we share the annotation guidelines and the annotated corpus.