Search

Abstract Background Clinical practice guidelines give recommendations about what to do in various medical situations, including therapeutical recommendations for drug prescription. An effective way to computerize these recommendations is to design critiquing decision support systems, i.e. systems that criticize the physician's prescription when it does not conform to the guidelines. These systems are commonly based on a list of "if conditions then criticism" rules. However, writing these rules from the guidelines is not a trivial task. The objective of this article is to propose methods that (1) simplify the implementation of guidelines' therapeutical recommendations in critiquing systems by automatically translating structured therapeutical recommendations into a list of "if conditions then criticize" rules, and (2) can generate an appropriate textual label to explain to the physician why his/her prescription is not recommended. Methods We worked on the therapeutic recommendations in five clinical practice guidelines concerning chronic diseases related to the management of cardiovascular risk. We evaluated the system using a test base of more than 2000 cases. Results Algorithms for automatically translating therapeutical recommendations into "if conditions then criticize" rules are presented. Eight generic recommendations are also proposed; they are guideline-independent, and can be used as default behaviour for handling various situations that are usually implicit in the guidelines, such as decreasing the dose of a poorly tolerated drug. Finally, we provide models and methods for generating a human-readable textual critique. The system was successfully evaluated on the test base. Conclusion We show that it is possible to criticize physicians' prescriptions starting from a structured clinical guideline, and to provide clear explanations. We are now planning a randomized clinical trial to evaluate the impact of the system on practices.

Background: Multidisciplinary tumor boards (MTBs) have been established in most countries to allow experts collaboratively determine the best treatment decisions for cancer patients. However, MTBs often face challenges such as case overload, which can compromise MTB decision quality. Clinical decision support systems (CDSSs) have been introduced to assist clinicians in this process. Despite their potential, CDSSs are still underutilized in routine practice. The emergence of large language models (LLMs), such as ChatGPT, offers new opportunities to improve the efficiency and usability of traditional clinical decision support systems (CDSSs)., Objectives: OncoDoc2 is a guideline-based CDSS developed using a documentary approach and applied to breast cancer management. This study aims to evaluate the potential of LLMs, used as question-answering (QA) systems, to improve the usability of OncoDoc2 across different prompt engineering techniques (PETs)., Methods: Data extracted from breast cancer patient summaries (BCPSs), together with questions formulated by OncoDoc2, were used to create prompts for various LLMs, and several PETs were designed and tested. Using a sample of 200 randomized BCPSs, LLMs and PETs were initially compared on their responses to OncoDoc2 questions using classic metrics (accuracy, precision, recall, and F1 score). Best performing LLMs and PETs were further assessed by comparing the therapeutic recommendations generated by OncoDoc2, based on LLM inputs, to those provided by MTB clinicians using OncoDoc2. Finally, the best performing method was validated using a new sample of 30 randomized BCPSs., Results: The combination of Mistral and OpenChat models under the enhanced zero-shot PET showed the best performance as a question-answering system. This approach gets a precision of 60.16%, a recall of 54.18%, an F1 Score of 56.59%, and an accuracy of 75.57% on the validation set of 30 BCPSs. However, this approach yielded poor results as a CDSS, with only 16.67% of the recommendations generated by OncoDoc2 based on LLM inputs matching the gold standard., Conclusions: All the criteria in the OncoDoc2 decision tree are crucial for capturing the uniqueness of each patient. Any deviation from a criterion alters the recommendations generated. Despite a good accuracy rate of 75.57% was achieved, LLMs still face challenges in reliably understanding complex medical contexts and be effective as CDSSs., Competing Interests: The authors declare that they have no conflict of interest., (The Author(s). This is an open access article published by Thieme under the terms of the Creative Commons Attribution License, permitting unrestricted use, distribution, and reproduction so long as the original work is properly cited. (https://creativecommons.org/licenses/by/4.0/).)

Clinical decision support systems for Nursing Process (NP-CDSSs) help resolve a critical challenge in nursing decision-making through automating the Nursing Process. NP-CDSSs are more effective when they are linked to Electronic Medical Record (EMR) Data allowing for the computation of Risk Assessment Scores. Braden scale (BS) is a well-known scale used to identify the risk of Hospital-Acquired Pressure Injuries (HAPIs). While BS is widely used, its specificity for identifying high-risk patients is limited. This study develops and evaluates a Machine Learning (ML) model to predict the HAPI risk, leveraging EMR readily available data. Various ML algorithms demonstrated superior performance compared to BS (pooled model AUC/F1-score of 0.85/0.8 vs. AUC of 0.63 for BS). Integrating ML into NP-CDSSs holds promise for enhancing nursing assessments and automating risk analyses even in hospitals with limited IT resources, aiming for better patient safety.

The healthcare sector confronts challenges from overloaded tumor board meetings, reduced discussion durations, and care quality concerns, necessitating innovative solutions. Integrating Clinical Decision Support Systems (CDSSs) has a potential in supporting clinicians to reduce the cancer burden, but CDSSs remain poorly used in clinical practice. The emergence of OpenAI's ChatGPT in 2022 has prompted the evaluation of Large Language Models (LLMs) as potential CDSSs for diagnosis and therapeutic management. We conducted a scoping review to evaluate the utility of LLMs like ChatGPT as CDSSs in several medical specialties, particularly in oncology, and compared users' perception of LLMs with the actually measured performance of these systems.

Text classification plays an essential role in the medical domain by organizing and categorizing vast amounts of textual data through machine learning (ML) and deep learning (DL). The adoption of Artificial Intelligence (AI) technologies in healthcare has raised concerns about the interpretability of AI models, often perceived as "black boxes." Explainable AI (XAI) techniques aim to mitigate this issue by elucidating AI model decision-making process. In this paper, we present a scoping review exploring the application of different XAI techniques in medical text classification, identifying two main types: model-specific and model-agnostic methods. Despite some positive feedback from developers, formal evaluations with medical end users of these techniques remain limited. The review highlights the necessity for further research in XAI to enhance trust and transparency in AI-driven decision-making processes in healthcare.

Using clinical decision support systems (CDSSs) for breast cancer management necessitates to extract relevant patient data from textual reports which is a complex task although efficiently achieved by machine learning but black box methods. We proposed a rule-based natural language processing (NLP) method to automate the translation of breast cancer patient summaries into structured patient profiles suitable for input into the guideline-based CDSS of the DESIREE project. Our method encompasses named entity recognition (NER), relation extraction and structured data extraction to systematically organize patient data. The method demonstrated strong alignment with treatment recommendations generated for manually created patient profiles (gold standard) with only 2% of differences. Moreover, the NER pipeline achieved an average F1-score of 0.9 across the main entities (patient, side, and tumor), of 0,87 for relation extraction, and 0.75 for contextual information, showing promising results for rule-based NLP.

The potential for technology to transform health care is tremendous, but advances in digital health may also bring privacy and data security challenges that may exacerbate inequalities. Hence, it is critical that the development of digital health is included in a framework of humanistic and ethical values. France drew up its roadmap for accelerating the shift towards digital health with ethics at the forefront, along with security and interoperability pillars. Criteria such as digital health for all, transparency of data processing, trustworthy AI, and eco-responsibility and sustainability of digital health were elaborated. Under the French Presidency of the Council of the European Union, building on the proposal of ethical criteria from France, eHealth network representatives unanimously adopted 16 European ethical principles for digital health, formalizing trust commitments towards European citizens and paving the way for the European Health Data Space.

The digital revolution is perceived as minimizing the environmental impact of human activities. However, digital actually uses real physical equipment, that consumes natural resources and emits greenhouse gases. The French Ministry of health has very soon engaged in the development of an ethical digital health with a strong concern for eco-responsibility. We have developed an Ecoscore service to allow providers of web and mobile health apps to calculate the environmental impact of their apps. The ecoscore is computed on a script describing the canonical use of the app. It is based on parameters actually measured, data flow (data downloaded from datacenters), performance (time taken to display the result of an action), and energy consumption (battery discharge rate). The French Ministry of Health has made the ecoscore mandatory to software providers candidating to the service catalog of Mon espace santé, a digital health space offered to all French citizens.

Breast cancer is the most commonly diagnosed cancer worldwide, and its burden has been rising over the past decades. A significant advance in healthcare is the integration of Clinical Decision Support Systems (CDSSs) into medical practice, which support healthcare professionals improving clinical decisions, leading to recommended patient-specific treatments and enhanced patient care. Breast cancer CDSSs are thus currently expanding, whether applied to screening, diagnostic, therapeutic or follow-up tasks. We conducted a scoping review to study their availability and use in practice. Except risk calculators, very few CDSSs are currently routinely used.

The search strategy of a literature review is of utmost importance as it impacts the validity of its findings. In order to build the best query to guide the literature search on clinical decision support systems applied to nursing clinical practice, we developed an iterative process capitalizing on previous systematic reviews published on similar topics. Three reviews were analyzed relatively to their detection performance. Errors in the choice of keywords and terms used in title and abstract (missing MeSH terms, failure to use common terms), may make relevant articles invisible.

Guideline-based clinical decision support systems (CDSSs) need the most recent evidence for reliable performance, making the provision of regularly updated clinical practice guidelines (CPGs) a major issue. Some international guidelines are renewed in short intervals and can be used for checking the status of given national guidelines with regard to the most recent evidence. Considering the volume of medical data and the number of CPGs published, computerized comparison of clinical guidelines can be an effective method. We performed a scoping review to evaluate the methods used for comparing two CPGs. We searched for methods for extracting CPG components and for methods used for comparing CPGs at different levels of abstraction. In each case, computerized and semi-computerized methods were recognized. Expert knowledge has yet a determinant role for assessing the comparisons, this role being more prominent for the extraction of semantic rules and the resolution of inconsistencies.

Although guaranteed by the GDPR, transparency of health data processing may not be fully respected, leading citizens to mistrust eHealth and discard digital health services. Identifying and safeguarding ethics in eHealth services is thus important to promote their development. We conducted a survey to assess the extent of ethical issues induced by the use of digital health services, understand the efforts citizens would be willing to accept for reporting such issues, and evaluate citizens' expectations regarding this reporting. Among 200 respondents, 36% reported having encountered ethical issues with the processing of their health data or with digital health services being poorly inclusive. Faced to ethical issues when using a digital health service, 49% of respondents were rather or very angry, and 33% felt rather or very dependent. Most respondents were ready to report digital health ethical issues if there is a feedback for each report.

In many countries, the management of cancer patients must be discussed in multidisciplinary tumor boards (MTBs). These meetings have been introduced to provide a collaborative and multidisciplinary approach to cancer care. However, the benefits of MTBs are now being challenged because there are a lot of cases and not enough time to discuss all the of them. During the evaluation of the guideline-based clinical decision support system (CDSS) of the DESIREE project, we found that for some clinical cases, the system did not produce recommendations. We assumed that these cases were complex clinical cases and needed deeper MTB discussions. In this work, we trained and tested several machine learning and deep learning algorithms on a labelled sample of 298 breast cancer patient summaries, to predict the complexity of a breast cancer clinical case. XGboost and multi-layer perceptron were the models with the best result, with an F1 score of 83%.

Using guideline-based clinical decision support systems (CDSSs) has improved clinical practice, especially during multidisciplinary tumour boards (MTBs) in cancer patient management. However, MTBs have been reported to be overcrowded, with limited time to discuss all cases. Complex breast cancer cases that need further MTB discussions should have priority in the organization of MTBs. In order to optimize MTB workflow, we attempted to predict complex cases defined as non-compliant cases despite the use of the decision support system OncoDoc. After previously obtaining insufficient performance with machine learning algorithms, we tested Multi Layer Perceptron for classification, compared various samplers to compensate data imbalance combined with cross- validation, and optimized all models with hyperparameter tuning and feature selection with no improvement and lacklustre results (F1-score: 31.4%).

The guideline-based decision support system (GL-DSS) of the DESIREE project and OncoDoc are two clinical decision support systems applied to the management of breast cancer. In order to evaluate the DESIREE GL-DSS, we decided to reuse a sample of clinical cases previously resolved by the multidisciplinary tumor board (MTB) of the Tenon Hospital (Paris, France) when using OncoDoc. Since we had two different knowledge representation models to represent clinical parameters and decisions, and two formalisms to represent guidelines, we developed a transformation sequence, involving the creation of synthetic patients, the enrichment of DESIREE ontology, and the translation of clinical cases and their decisions, to transform OncoDoc data into the DESIREE representation. Considering MTB decisions as the gold standard, the 84% compliance rate of DESIREE recommendations was rather satisfactory. Some situations (0.7%) concerned clinical cases that were compliant neither with OncoDoc nor with DESIREE that we defined as complex cases, not handled by guidelines, which necessitate effective MTB discussions.

Computerized decision support systems (CDSSs) are still poorly routinely implemented in clinical practices mainly because of usability problems related to the technology interface. We previously proposed to use gauges to visualize the output of a guideline-based CDSS applied to malnutrition and pressure ulcer management in nursing homes. This interface was assessed by four focus groups including 16 healthcare professionals with expertise in geriatrics. A USE-like questionnaire was distributed. Participants considered the dashboard-with-gauges visualization was useful (94%), easy to use (63%), easy to learn (88%), and 88% thought they could be satisfied with it. However, concerns were expressed about the difficulty to follow up multiple healthcare problems.

Medication reconciliation (MR) aims at preventing medication errors at care transitions. It is a complex, time-consuming, cognitively demanding pharmacological task. We have developed a decision support system, EzMedRec, to assist retroactive MR at hospital admission. EzMedRec compares the best possible medication history (BPMH), i.e., all medications taken by the patient before hospitalization, to the list of admission medication orders (AMO). The process includes (i) the decomposition of BPMH and AMO drugs into their active ingredients (AIs), (ii) the detection of medication discontinuations and additions, and (iii) the identification of modified medication orders. The ATC classification is used to semantically enrich MR by comparing discontinued AIs and added AIs and suggesting a potential intentional drug substitution serving the same therapeutic objective. EzMedRec has been evaluated on a sample of 52 actual MRs involving 822 medication order lines, 406 in BPMHs, and 416 in AMOs with a global accuracy of 98,3%., (©2020 AMIA - All rights reserved.)

Though a preventable risk, the management of pressure ulcers (PUs) in nursing homes is not satisfactory due to inadequate prevention and complex care plans. PUs early detection and wound assessment require to know the patient condition and risk factors and to have a good knowledge of best practices. We built a guideline-based clinical decision support system (CDSS) for the prevention, the assessment, and the management of PUs. Clinical practice guidelines have been modeled as decision trees and formalized as IF-THEN rules to be triggered by electronic health record (EHR) data. From PU assessment yielded by the CDSS, we propose a synthetic visualization of PU current and previous stages as a gauge that illustrates the different stages of PU continuous evolution. This allows to display PU current and previous stages to inform health care professionals of PU updated assessment and support their evaluation of previously delivered care efficiency. The CDSS will be integrated in NETSoins nursing homes EHR where gauges for several health problems constitute a patient dashboard.

Interoperability issues are common in biomedical informatics. Reusing data generated from a system in another system, or integrating an existing clinical decision support system (CDSS) in a new organization is a complex task due to recurrent problems of concept mapping and alignment. The GL-DSS of the DESIREE project is a guideline-based CDSS to support the management of breast cancer patients. The knowledge base is formalized as an ontology and decision rules. OncoDoc is another CDSS applied to breast cancer management. The knowledge base is structured as a decision tree. OncoDoc has been routinely used by the multidisciplinary tumor board physicians of the Tenon Hospital (Paris, France) for three years leading to the resolution of 1,861 exploitable decisions. Because we were lacking patient data to assess the DESIREE GL-DSS, we investigated the option of reusing OncoDoc patient data. Taking into account that we have two CDSSs with two formalisms to represent clinical practice guidelines and two knowledge representation models, we had to face semantic and structural interoperability issues. This paper reports how we created 10,681 synthetic patients to solve these issues and make OncoDoc data re-usable by the GL-DSS of DESIREE.

Nursing home (NH) residents are known to be at risk of preventable adverse events due to inadequate monitoring or failure to provide necessary treatments. Missed care has been partially explained by nurses' lack of knowledge. We describe a guideline-based decision support system for the management of malnutrition in NHs. Three steps are distinguished: screening, therapeutic management, and follow-up. Clinical practice guidelines have been modeled as decision trees and formalized as IF-THEN rules to be triggered by electronic health records data (e.g., weight, albuminemia). We propose a visualization of recommendations as a dashboard of gauges displaying both current and previous levels of malnutrition to inform on the effect of therapeutic management and facilitate a correct follow-up.

Clinical practice guidelines (CPGs) often include ambiguous criteria making their translation as computer-interpretable guidelines a difficult task. In breast cancer management, whether to perform a breast conservative surgery (BCS) or not is one example. Most international CPGs recommend to perform a BCS when the tumour volume / breast volume ratio allows for good cosmetic results, which cannot be directly translated into a computable format. We propose to compute an estimate of the ratio using the maximum size of the tumour to compute the tumour volume and the bra size to compute the breast volume. In addition, we take into account the location of the tumour according to quadrants and unions of quadrants. The model has been tested on a retrospective sample of 34 clinical decisions of a breast cancer unit in a Parisian university hospital (France). Concordance was found in 91.2% of the cases, with good sensibility and specificity. This finding could set a new pathway to advance on the development of actionable decision criteria to be used in a future clinical decision support system for breast cancer management.

The world population is dramatically ageing, resulting in an increase of the prevalence of older dependent adults living in nursing homes (NHs). Because of insufficient resources in NHs, and nurses' lack of time and knowledge, adverse events, most of them being preventable, are often reported. Clinical decision support systems (CDSSs) have proven to improve the quality of care in various healthcare settings such as hospitals and primary care centers. However, the use of CDSSs in NHs is still limited and little is known about their influence on nursing practices and NH residents' clinical outcomes. We conducted a scoping review of the literature to evaluate CDSS use in NHs. Out of 1,231 retrieved papers, 15 studies were retrieved which assess 10 CDSSs applied to pressure ulcers and malnutrition prevention, drug prescription, and disease management. This review showed CDSSs could be effective in NHs for improving routine clinical practice and patient outcomes, but research is still needed to implement effective CDSSs in NHs.

The International Medical Informatics Association (IMIA), a non-governmental, not-for-profit, global organization promoting health and biomedical informatics, is committed to the right of communities/populations and individuals to science, comprised of three separate constituent rights: 1) the right to participate in science, 2) the right to benefit from science, and 3) the right to benefit from a person's own contribution to science or inventions. As such, IMIA provides a global platform where scientists, researchers, health information users, vendors, developers, consultants, health care consumers, and suppliers can meet in an environment of cooperation and sharing. In the context of IMIA's conferences, the IMIA board has discussed and identified the important central factors, which are essential considerations to host a scientific meeting. These factors will be used to help vet future contenders applying for the honor to host an IMIA conference: Reasonable safety and security, commitment by the host member society, freedom of travel, scientific freedom, and freedom from discrimination., Competing Interests: Disclosure The authors report no conflicts of interest in this work., (Georg Thieme Verlag KG Stuttgart.)

Clinical Practice Guidelines (CPGs) gather latest evidence-based results to guide and support clinicians over the decision-making process to provide best care. Nevertheless, clinical cases may be subject to some biases (understood as non-compliance with CPGs) that can lead to adapt care delivery. In this work an experience-based decision support leaning on the structuration of the Decisional Event concept for tracking and storing each clinical decision is presented. Moreover, a visual analytics tool is provided in order to facilitate the visualization of biases from guideline-based decision support and the identification and inclusion of real-world evidence into the reasoning process by augmenting the knowledge formalized in the implemented guidelines.