Healthcare Decision Support Research Articles

Multi-criteria and medical diagnosis for application to health insurance systems: a general approach through non-additive measures

In this contribution, we propose a healthcare decision support system. Nowadays, it is commonly recognized that quantitative tools and decision support can increase the benefits and the performances of healthcare systems, and for this, different multiple criteria methods were proposed in many branches of medicine. The approach we propose is based on non-additive measures and the Choquet integral. This methodology has been intensively applied in many real-world applications, given its capability to represent interactions among criteria, and thus to model a wide range of preference structures. Considering that the diagnosis procedure needs also to take the clinical expertise into account, this method appears particularly tailored for a diagnosis support, mainly when statistical models cannot be applied and/or available data are scarce and knowledge can be inferred by physicians’ opinions. In particular, we propose a disease risk evaluation and compute some associated indicators. Furthermore, an error estimation is performed. As an application, a cardiovascular risk diagnosis model is presented. The proposed methodology, that allows to quantify the disease risk taking into account individual’s medical conditions, can be used for improving healthcare service quality or for pricing and reserving health insurance policies. An application to health insurance pricing is provided.

Abstract 290: Lessons Learned From Integrated Management Program Advancing Community Treatment Of Atrial Fibrillation (IMPACT-AF): A Pragmatic Clinical Trial Of Computerized Decision Support In Primary Healthcare

Background: Integrated Management Program Advancing Community Treatment of Atrial Fibrillation (IMPACT-AF) was a pragmatic, cluster randomized trial assessing the clinical relevance and effectiveness of a clinical decision support (CDS) tool in the primary care setting of Nova Scotia, Canada. Key challenges encountered included CDS development and implementation (2013-2018), study recruitment (2014-2016) and data analysis (2018 to present). Methods: Clinical and health informatics researchers developed CDS software designed to help primary care providers (PCP) deliver individualized AF patient care based on national guidelines. Features included prioritized automated alerts signaling material changes in patient clinical or biochemical profiles requiring expedited treatment changes. Challenges documented over the trial duration are presented here as lessons learned. Findings: 1) Resources must be allocated for feasibility testing and software development/updates. CDS development took twice as long as projected. Network access and broadband speeds were key impediments to successful uptake. Although modified after pilot testing with an initial cohort of intervention providers, user feedback at study completion suggested the CDS was not sufficiently user-friendly and did not create efficiencies in the clinical management of AF for patients (i.e., repetitive alerts). 2) Integration across e-platforms is crucial. Intellectual property and other technical issues prohibited integration of the CDS within providers’ existing electronic medical records and desired provincial e-health platforms. PCPs cited double data entry/login as impediments to participation/reasons for withdraw. Challenges with data integration across platforms prevented facile and timely data access, analysis and reporting. 3) Study recruitment is resource intensive. In total, 204 PCPs and 1,204 patients participated, representing 25% of all eligible PCPs and 13% of all persons living with AF in Nova Scotia, respectively. The most effective PCP recruitment strategy was in-office, small group lunch and learns. PCPs with past research experience/those who led patient consent processes were top patient recruiters. The study office played a pivotal role in achieving recruitment targets. Conclusions: A rapid growth in healthcare data is leading to the widespread development of CDS software to analyse it. Our experience found practical issues to address if such applications are to succeed. CDS tools that fully integrate multiple critical co-morbid guideline recommendations across eHealth platforms should be pursued. Feasibility testing to assess the practical utility of any healthcare CDS software prior to its implementation is recommended. Lastly, adequate resources are necessary to support successful recruitment of PCPs for future pragmatic clinical trials.

A novel healthcare resource allocation decision support tool: A forecasting-simulation-optimization approach

The increasing pressures on the healthcare system in the UK are well documented. The solution lies in making best use of existing resources (e.g. beds), as additional funding is not available. Increasing demand and capacity shortages are experienced across all specialties and services in hospitals. Modelling at this level of detail is a necessity, as all the services are interconnected, and cannot be assumed to be independent of each other. Our review of the literature revealed two facts; First an entire hospital model is rare, and second, use of multiple OR techniques are applied more frequently in recent years. Hybrid models which combine forecasting, simulation and optimization are becoming more popular. We developed a model that linked each and every service and specialty including A&E, and outpatient and inpatient services, with the aim of, (1) forecasting demand for all the specialties, (2) capturing all the uncertainties of patient pathway within a hospital setting using discrete event simulation, and (3) developing a linear optimization model to estimate the required bed capacity and staff needs of a mid-size hospital in England (using essential outputs from simulation). These results will bring a different perspective to key decision makers with a decision support tool for short and long term strategic planning to make rational and realistic plans, and highlight the benefits of hybrid models.

Improving implantable cardioverter defibrillator deactivation discussions in admitted patients made DNR and comfort care

BackgroundUnintended shocks from implantable cardioverter defibrillators (ICDs) are often distressing to patients and family members, particularly at the end of life. Unfortunately, a large proportion of ICDs remain active at...

Reference care plans and clinical pathways, from standardisation to personalisation

Thereference care plan and clinical pathway are complementary tools. Their implementation, once the method is acquired, is a real scientific approach. It formalises a clinical reasoning shared between the professionals involved in the patient's care. These tools are quality reference guides which give meaning to the caregiving activity. They help to consolidate competencies. This approach guides the clinical judgement and reasoning of caregivers. It is a health care decision support tool. This collective work is the result of reflective practice and is built on experiential knowledge.

Multi-agent based intuitionistic fuzzy logic healthcare decision support system

Multi-agent based intuitionistic fuzzy logic healthcare decision support system

A data integration platform for patient-centered e-healthcare and clinical decision support

The proliferation of disconnected information systems accumulating increasing volumes of heterogeneous data is a formidable challenge in the healthcare domain. It is necessary to address this challenge as the entire domain gradually transitions into e-healthcare and patient-centered healthcare delivery models. The lack of data integration at systems level prohibits clinicians from building a comprehensive understanding of the patient condition and further deprives patients of developing a holistic awareness of their medical conditions. In this paper, we propose an open data integration platform for patient, clinical, medical and historical data siloed across multiple health information systems. As an open platform, it can accommodate and integrate further heterogeneous data sources such as data streams generated by wearable IoT devices. As an integration platform, it facilitates the centralization of data assets. This centralization empowers every stakeholder in a patient-centered care setting to actively participate in decision-making. A range of analytics and reporting solutions, such as data warehouse, interactive dashboards, and predictive analytics tools, can be deployed upon this open data integration platform. The proposed platform is currently being adapted and implemented to address patient-centered healthcare and clinical decision support requirements in a sports injury clinic at a not-for-profit private hospital in Melbourne, Australia. Use Case based demonstration of the platform’s suitability for holistic information management, decision support, and predictive analytics justify its role in the advancement of e-healthcare.

Fujitsu HIKARI, a Healthcare Decision Support System based on Biomedical Knowledge

Fujitsu HIKARI is an artificial intelligence solution to assist clinicians in medical decision making, developed in the context of a joint collaboration project between Fujitsu Laboratories of Europe and Hospital Clínico San Carlos. This decision support system leverages on data analytics combined with healthcare semantic information to provide health estimations for patients, improving care quality and personalized treatment. Fujitsu HIKARI stands on the shoulders of biomedical knowledge, which includes (i) theoretical knowledge extracted from scientific literature, domain expert knowledge, and health standards; and (ii) empirical knowledge extracted from real patient electronic health records. The theoretical knowledge combines a theoretical knowledge graph (TKG) and a biomedical document repository (BDR). The empirical knowledge is encoded in an empirical knowledge graph (EKG). One of the main functionalities of Fujitsu HIKARI is the patient mental health risks assessment, which is based on the exploitation of its underlying Biomedical Knowledge.

‘MEAN+R’: implementing a web-based, multi-participant decision support system using the prevalent MEAN architecture with R based on a revised intuitionistic-fuzzy multiple attribute decision-making model

This study proposes the novel idea of implementing a decision support system (DSS) based on both the prevalent MEAN architecture and the R statistical language platform. A ‘MEAN+R’ framework is defined, and the effectiveness of this framework is verified by taking it as the basis for the implementation of a real user-side, web-based, multi-participant DSS supporting a group decision-making process of quasi-senior people, pertaining to the selection of the best and commonly-agreed senior center for co-living after retirement. The ‘model base’ of the DSS is a recent MADM model evaluating the criteria-wise source opinions of the decision makers (DMs) toward alternatives in terms of intuitionistic fuzzy numbers (IFNs), but within the model the original multi-objective decision-making (MODM) phase to determine whether the criteria weight vector (CWV) has been replaced with another well-accepted MADM model, which is the AHP. This ensures not only unified grounds for modelling, but also a coherent implementation basis for the MADM model in R. The outside Node.js programs access the ‘data base’ of the DSS managed using MongoDB and will invoke the R-written ‘model base’ codes while necessary. Then, the R program also interacts with the ‘data base’ to exchange the information relevant to the decision from/to users. An illustrative numerical group decision example supported by this DSS is provided, and an efficiency analysis is made by listing the time spent on both data and model bases separately and clearly, while excluding the other processes of the DSS. A key finding is that the computational time can be controlled when incorporating the vector-based R platform, i.e., it leaps forward to the next level only when the size of the problem increases to a certain extent. Finally, since the realized DSS is intended to face end users of the healthcare system (rather than the business-oriented service providers, e.g., hospitals, physicians, nurses, etc.), it may provide some supplement to healthcare decision support (rather than solving other types of problem, such as clinical decisions).

A preliminary psychometric evaluation of the eight-item cognitive load scale

AimThe aim of this article is to report the psychometric properties of the eight-item cognitive load scale. BackgroundAccording to cognitive load theory, the formatting and delivery of healthcare education influences the degree to which patients and/or family members can engage their working memory systems for learning. However, despite its relevance, cognitive load has not yet been evaluated among surrogate decision makers exposed to electronic decision support for healthcare decisions. To date, no psychometric analyses of instruments evaluating cognitive load have been reported within healthcare settings. MethodsA convenience sample of 62 surrogate decision makers for critically ill patients were exposed to one of two healthcare decision support interventions were recruited from four intensive care units at a tertiary medical center in Northeast Ohio. Participants were administered a battery of psychosocial instruments and the eight-item cognitive load scale (CLS). ResultsThe CLS demonstrated a bidimensional factor structure with acceptable discriminant validity and internal consistency reliability (Cronbach's α = 0.75 and 0.89). ConclusionsThe CLS is a psychometrically sound instrument that may be used in the evaluation of decision support among surrogate decision makers of the critically ill. The authors recommend application of the cognitive load scale in the evaluation and development of healthcare education and interventions.

Healthcare decision support tool: multi-agent system for bed management

Healthcare decision support tool: multi-agent system for bed management

Healthcare decision support tool: multi-agent system for bed management

The effective management of the allocation of limited resources such as beds is a key to the delivery of timely, high quality and low cost healthcare. However, an efficient utilisation of beds is not easy to accomplish because of the complexity of the healthcare system and the higher amount of problems that the hospitals are facing such as the overcrowding. Moreover, in order to solve these problems, effective bed management remains one of the prominent solutions for the majority of hospitals. In this paper, a decision support system based on multi-agent systems has been proposed. The suggested tool will assist healthcare practitioners to manage beds in the hospitals and also to address the problem of bed management by offering a referral depending on availability of beds to avoid tension in hospitals.

Prevalence of antibiotic prescription in pediatric outpatients in Italy: the role of local health districts and primary care physicians in determining variation. A multilevel design for healthcare decision support

BackgroundAccording to scientific literature, antibacterials are prescribed for common pediatric conditions that do not benefit from antibiotic therapy. The link between antibiotic use and bacterial resistance is well known. Antibiotic overprescribing generates high social costs and severe consequences for children. Our objectives were to analyze antibiotic prescription patterns in pediatric outpatients residing in the Lazio region (Italy), to identify physicians’ characteristics associated with increased propensity for prescription, to identify the priority axes for action to improve the rational use of antibiotics.MethodsWe enrolled all children aged 13 years or less in 2014. Antibiotic prescription patterns were analyzed during a one-year follow-up period. The main outcome measures were the antibiotic prescription prevalence, and the geographic variation in antibiotic prescribing. Multilevel models were performed to analyze variation. Variation was expressed as Median Odds Ratios (MORs). If the MOR is 1.00, there is no variation between clusters. If there is considerable between-cluster variation, the MOR will be large.ResultsWe enrolled 636,911 children. Most of them were aged 6–13 years (57.3%). In 2015, the antibiotic prescription prevalence was 46% in the 0–13, 58% in the 0–5, and 37% in the 6–13 age group. Overall, penicillins were the most prescribed antibiotics, their consumption increased from 43% to 52% during the 2007–2015 period. In 2015, the antibiotic prescription prevalence ranged from 30% to 62% across local health districts (LHDs) of the region. Moreover, a significant (p < 0.001) variation was observed between physicians working in the same LHD: MORs were equal to 1.52 (1.48–1.56) and 1.46 (1.44–1.48) in the 0–5 and 6–13 age groups, respectively. The probability of prescribing antibiotics was significantly (p < 0.001) lower for more-experienced physicians.ConclusionsPediatric antibiotic use in the Lazio region is much higher than in other European countries. The intra-regional drug prescribing variability underlines the lack of therapeutic protocols shared at regional level and raises equity issues in access to optimal care. Both LHD managers and individual physicians should be involved in training interventions to improve the targeted use of antibiotics and mitigate the effect of contextual variables, such as the spatial-related socioeconomic status of the patient/parent binomial.

Healthcare support for underserved communities using a mobile social media platform

Emerging digital technologies for healthcare information support have already contributed to reducing the digital divide among rural communities. Although mobile health (m-health) applications facilitate provision of support for treatment consultation in real-time, their substantial potential has not yet been operationalised for decision support to meet citizen demand in developing nations. Modern healthcare information access, especially in rural areas of developing countries, is critical to effective healthcare, since both information and expert opinions are limited. Mobile phone and social media penetration, however, is often extensive. In this paper, we design and evaluate an innovative mobile decision support system (MDSS) solution for rural citizens healthcare decision support and information dissemination.Developed using a design science approach, the instantiated artifact connects underserved rural patients in Bangladesh to general practitioners (GPs) – allowing GPs, based on queries and information support provided, to evaluate patient conditions virtually and provide answers for further diagnosis or treatment. A cloud platform using social media embodies health record information and is used with a rating technique that matches queries to profiled remote experts, participating asynchronously. A comprehensive evaluation of the MDSS artifact ensures its utility, efficacy, and reliability.

Approach of HeDSS (Health-care Decision Support System) using Context-aware Technique

Approach of HeDSS (Health-care Decision Support System) using Context-aware Technique

Designing the Right Framework for Healthcare Decision Support

Many factors need to be taken into consideration while developing a decision support system for healthcare. This mainly involves: a) adherence to statutory regulations, b) ease of use and access, and c) protecting patient data from malicious use. Some of these requirements are intertwined creating a myriad of complexities. This leads to a substantial increase in the level of complexity involved in designing and developing the decision support system. In this paper we attempt to address some of these complexities to the reader and present a framework for a solution that could be modified if required to deal with these aforementioned complexities.

Decision support in healthcare: determining provider influence on treatment outcomes with robust risk adjustment

Recent advances in decision support offer a powerful toolkit to monitor, benchmark and improve healthcare quality. Since the underlying demography and prior illness of patients vary across hospitals, performance should be risk-adjusted to allow for fair comparison of providers. While previous research has proposed risk-standardisation models, their estimation still remains a non-trivial task. As a remedy, this paper utilises the least absolute shrinkage operator, LASSO for short, as a Bayesian method as part of a framework to propose a robust technique for estimating the risk-adjusted performance of hospitals. We compare our robust estimation to existing approaches (namely, the case mix approach and hierarchical generalised linear models) and demonstrate its advantages with real data. Our findings show that our method, which is robust and fully automated, can yield useful findings even with small data sets.

Forecasting influenza outbreak dynamics in Melbourne from Internet search query surveillance data.

BackgroundAccurate forecasting of seasonal influenza epidemics is of great concern to healthcare providers in temperate climates, as these epidemics vary substantially in their size, timing and duration from year to year, making it a challenge to deliver timely and proportionate responses. Previous studies have shown that Bayesian estimation techniques can accurately predict when an influenza epidemic will peak many weeks in advance, using existing surveillance data, but these methods must be tailored both to the target population and to the surveillance system.ObjectivesOur aim was to evaluate whether forecasts of similar accuracy could be obtained for metropolitan Melbourne (Australia).MethodsWe used the bootstrap particle filter and a mechanistic infection model to generate epidemic forecasts for metropolitan Melbourne (Australia) from weekly Internet search query surveillance data reported by Google Flu Trends for 2006–14.Results and ConclusionsOptimal observation models were selected from hundreds of candidates using a novel approach that treats forecasts akin to receiver operating characteristic (ROC) curves. We show that the timing of the epidemic peak can be accurately predicted 4–6 weeks in advance, but that the magnitude of the epidemic peak and the overall burden are much harder to predict. We then discuss how the infection and observation models and the filtering process may be refined to improve forecast robustness, thereby improving the utility of these methods for healthcare decision support.

Analysis of Data Mining Techniques for Healthcare Decision Support System Using Liver Disorder Dataset

Accuracy in data classification depends on the dataset used for learning. Now-a-days the most important cause of death for both men and women is due to the Liver Problem. The healthcare industry collects a huge amount of data which is not properly mined and not put to the optimum use. Discovery of these hidden patterns and relationships often goes unexploited. Our research focuses on this aspect of Medical diagnosis by learning pattern through the collected data of Liver disorder to develop intelligent medical decision support systems to help the physicians. In this paper, we propose the use decision trees J48, Naive Bayes, ANN, ZeroR, 1BK and VFI algorithm to classify these diseases and compare the effectiveness, correction rate among them. Detection of Liver disease in its early stage is the key of its cure. It leads to better performance of the classification models in terms of their predictive or descriptive accuracy, diminishing of computing time needed to build models as they learn faster, and better understanding of the models. In this paper, a comparative analysis of data classification accuracy using Liver disorder data in different scenarios is presented. The predictive performances of popular classifiers are compared quantitatively.

Methodological Challenges in Examining the Impact of Healthcare Predictive Analytics on Nursing-Sensitive Patient Outcomes.

The expansion of real-time analytic abilities within current electronic health records has led to innovations in predictive modeling and clinical decision support systems. However, the ability of these systems to influence patient outcomes is currently unknown. Even though nurses are the largest profession within the healthcare workforce, little research has been performed to explore the impact of clinical decision support on their decisions and the patient outcomes associated with them. A scoping literature review explored the impact clinical decision support systems containing healthcare predictive analytics have on four nursing-sensitive patient outcomes (pressure ulcers, failure to rescue, falls, and infections). While many articles discussed variable selection and predictive model development/validation, only four articles examined the impact on patient outcomes. The novelty of predictive analytics and the inherent methodological challenges in studying clinical decision support impact are likely responsible for this paucity of literature. Major methodological challenges include (1) multilevel nature of intervention, (2) treatment fidelity, and (3) adequacy of clinicians' subsequent behavior. There is currently insufficient evidence to demonstrate efficacy of healthcare predictive analytics-enhanced clinical decision support systems on nursing-sensitive patient outcomes. Innovative research methods and a greater emphasis on studying this phenomenon are needed.