Search

Rationale: This paper aims to show how the focus on eradicating bias from Machine Learning decision-support systems in medical diagnosis diverts attention from the hermeneutic nature of medical decision-making and the productive role of bias. We want to show how an introduction of Machine Learning systems alters the diagnostic process. Reviewing the negative conception of bias and incorporating the mediating role of Machine Learning systems in the medical diagnosis are essential for an encompassing, critical and informed medical decision-making., Methods: This paper presents a philosophical analysis, employing the conceptual frameworks of hermeneutics and technological mediation, while drawing on the case of Machine Learning algorithms assisting doctors in diagnosis. This paper unravels the non-neutral role of algorithms in the doctor's decision-making and points to the dialogical nature of interaction not only with the patients but also with the technologies that co-shape the diagnosis., Findings: Following the hermeneutical model of medical diagnosis, we review the notion of bias to show how it is an inalienable and productive part of diagnosis. We show how Machine Learning biases join the human ones to actively shape the diagnostic process, simultaneously expanding and narrowing medical attention, highlighting certain aspects, while disclosing others, thus mediating medical perceptions and actions. Based on that, we demonstrate how doctors can take Machine Learning systems on board for an enhanced medical diagnosis, while being aware of their non-neutral role., Conclusions: We show that Machine Learning systems join doctors and patients in co-designing a triad of medical diagnosis. We highlight that it is imperative to examine the hermeneutic role of the Machine Learning systems. Additionally, we suggest including not only the patient, but also colleagues to ensure an encompassing diagnostic process, to respect its inherently hermeneutic nature and to work productively with the existing human and machine biases., (© 2021 The Authors. Journal of Evaluation in Clinical Practice published by John Wiley & Sons Ltd.)

This paper introduces a special section of the current issue of the Journal of Evaluation in Clinical Practice that includes a set of 6 empirical articles showcasing a versatile, new machine‐learning statistical method, known as optimal data (or discriminant) analysis (ODA), specifically designed to produce statistical models that maximize predictive accuracy. As this set of papers clearly illustrates, ODA offers numerous important advantages over traditional statistical methods—advantages that enhance the validity and reproducibility of statistical conclusions in empirical research. This issue of the journal also includes a review of a recently published book that provides a comprehensive introduction to the logic, theory, and application of ODA in empirical research. It is argued that researchers have much to gain by using ODA to analyze their data. [ABSTRACT FROM AUTHOR]

Rationale, Aims and Objectives: Interventions with multivalued treatments are common in medical and health research; examples include comparing the efficacy of competing interventions and contrasting various doses of a drug. In recent years, there has been growing interest in the development of methods that estimate multivalued treatment effects using observational data. This paper extends a previously described analytic framework for evaluating binary treatments to studies involving multivalued treatments utilizing a machine learning algorithm called optimal discriminant analysis (ODA)., Method: We describe the differences between regression-based treatment effect estimators and effects estimated using the ODA framework. We then present an empirical example using data from an intervention including three study groups to compare corresponding effects., Results: The regression-based estimators produced statistically significant mean differences between the two intervention groups, and between one of the treatment groups and controls. In contrast, ODA was unable to discriminate between distributions of any of the three study groups., Conclusions: Optimal discriminant analysis offers an appealing alternative to conventional regression-based models for estimating effects in multivalued treatment studies because of its insensitivity to skewed data and use of accuracy measures applicable to all prognostic analyses. If these analytic approaches produce consistent treatment effect P values, this bolsters confidence in the validity of the results. If the approaches produce conflicting treatment effect P values, as they do in our empirical example, the investigator should consider the ODA-derived estimates to be most robust, given that ODA uses permutation P values that require no distributional assumptions and are thus, always valid., (© 2016 John Wiley & Sons, Ltd.)

Rationale, Aims and Objectives: Program evaluations often utilize various matching approaches to emulate the randomization process for group assignment in experimental studies. Typically, the matching strategy is implemented, and then covariate balance is assessed before estimating treatment effects. This paper introduces a novel analytic framework utilizing a machine learning algorithm called optimal discriminant analysis (ODA) for assessing covariate balance and estimating treatment effects, once the matching strategy has been implemented. This framework holds several key advantages over the conventional approach: application to any variable metric and number of groups; insensitivity to skewed data or outliers; and use of accuracy measures applicable to all prognostic analyses. Moreover, ODA accepts analytic weights, thereby extending the methodology to any study design where weights are used for covariate adjustment or more precise (differential) outcome measurement., Method: One-to-one matching on the propensity score was used as the matching strategy. Covariate balance was assessed using standardized difference in means (conventional approach) and measures of classification accuracy (ODA). Treatment effects were estimated using ordinary least squares regression and ODA., Results: Using empirical data, ODA produced results highly consistent with those obtained via the conventional methodology for assessing covariate balance and estimating treatment effects., Conclusions: When ODA is combined with matching techniques within a treatment effects framework, the results are consistent with conventional approaches. However, given that it provides additional dimensions and robustness to the analysis versus what can currently be achieved using conventional approaches, ODA offers an appealing alternative., (© 2016 John Wiley & Sons, Ltd.)

Abstract: Rationale, aims, and objectives: Mediation analysis identifies causal pathways by testing the relationships between the treatment, the outcome, and an intermediate variable that mediates the relationship between the treatment and outcome. This paper introduces classification tree analysis (CTA), a machine‐learning procedure, as an alternative to conventional methods for analysing mediation effects. Method: Using data from the JOBS II study, we compare CTA to structural equation models (SEMs) by assessing their consistency in revealing mediation effects on 2 outcomes; reemployment (a binary variable) and depressive symptoms (a continuous variable). Because study participants were not randomized sequentially to both treatment and mediator, an additional model was generated including baseline covariates to strengthen the validity of some key identifying assumptions required of all mediation analyses. Results: Using SEM, no statistically significant treatment or mediated effects were found for either outcome. In contrast, CTA found a significant treatment effect for reemployment (P = .047) and a mediated pathway for individuals in the treatment group (P = .014). No CTA model could be generated for the reemployment outcome. When covariates were added to the model, CTA found numerous interactions, while SEM found no effects. Conclusions: CTA may uncover mediation effects where conventional approaches do not, because CTA does not require any assumptions about the distribution of variables nor of the functional form of the model, and CTA will systematically identify all statistically viable interactions. The versatility of CTA enables the investigator to explore the theorized underlying causal mechanism of an intervention in a much more comprehensive manner than conventional mediation analytic approaches. [ABSTRACT FROM AUTHOR]

Data mining techniques are gaining in popularity among health researchers for an array of purposes, such as improving diagnostic accuracy, identifying high‐risk patients and extracting concepts from unstructured data. In this paper, we describe how these techniques can be applied to another area in the health research domain: identifying characteristics of individuals who do and do not choose to participate in observational studies. In contrast to randomized studies where individuals have no control over their treatment assignment, participants in observational studies self‐select into the treatment arm and therefore have the potential to differ in their characteristics from those who elect not to participate. These differences may explain part, or all, of the difference in the observed outcome, making it crucial to assess whether there is differential participation based on observed characteristics. As compared to traditional approaches to this assessment, data mining offers a more precise understanding of these differences. To describe and illustrate the application of data mining in this domain, we use data from a primary care‐based medical home pilot programme and compare the performance of commonly used classification approaches – logistic regression, support vector machines, random forests and classification tree analysis (CTA) – in correctly classifying participants and non‐participants. We find that CTA is substantially more accurate than the other models. Moreover, unlike the other models, CTA offers transparency in its computational approach, ease of interpretation via the decision rules produced and provides statistical results familiar to health researchers. Beyond their application to research, data mining techniques could help administrators to identify new candidates for participation who may most benefit from the intervention. [ABSTRACT FROM AUTHOR]

Rationale, aims and objectives: Single‐group interrupted time series analysis (ITSA) is a popular evaluation methodology in which a single unit of observation is being studied, the outcome variable is serially ordered as a time series and the intervention is expected to ‘interrupt’ the level and/or trend of the time series, subsequent to its introduction. Given that the internal validity of the design rests on the premise that the interruption in the time series is associated with the introduction of the treatment, treatment effects may seem less plausible if a parallel trend already exists in the time series prior to the actual intervention. Thus, sensitivity analyses should focus on detecting structural breaks in the time series before the intervention. Method: In this paper, we introduce a machine‐learning algorithm called optimal discriminant analysis (ODA) as an approach to determine if structural breaks can be identified in years prior to the initiation of the intervention, using data from California's 1988 voter‐initiated Proposition 99 to reduce smoking rates. Results: The ODA analysis indicates that numerous structural breaks occurred prior to the actual initiation of Proposition 99 in 1989, including perfect structural breaks in 1983 and 1985, thereby casting doubt on the validity of treatment effects estimated for the actual intervention when using a single‐group ITSA design. Conclusions: Given the widespread use of ITSA for evaluating observational data and the increasing use of machine‐learning techniques in traditional research, we recommend that structural break sensitivity analysis is routinely incorporated in all research using the single‐group ITSA design. [ABSTRACT FROM AUTHOR]

In order to assess the effectiveness of matching approaches in observational studies, investigators typically present summary statistics for each observed pre‐intervention covariate, with the objective of showing that matching reduces the difference in means (or proportions) between groups to as close to zero as possible. In this paper, we introduce a new approach to distinguish between study groups based on their distributions of the covariates using a machine‐learning algorithm called optimal discriminant analysis (ODA). Assessing covariate balance using ODA as compared with the conventional method has several key advantages: the ability to ascertain how individuals self‐select based on optimal (maximum‐accuracy) cut‐points on the covariates; the application to any variable metric and number of groups; its insensitivity to skewed data or outliers; and the use of accuracy measures that can be widely applied to all analyses. Moreover, ODA accepts analytic weights, thereby extending the assessment of covariate balance to any study design where weights are used for covariate adjustment. By comparing the two approaches using empirical data, we are able to demonstrate that using measures of classification accuracy as balance diagnostics produces highly consistent results to those obtained via the conventional approach (in our matched‐pairs example, ODA revealed a weak statistically significant relationship not detected by the conventional approach). Thus, investigators should consider ODA as a robust complement, or perhaps alternative, to the conventional approach for assessing covariate balance in matching studies. [ABSTRACT FROM AUTHOR]

Rationale, aims and objectives: Establishing the relationship between various doses of an exposure and a response variable is integral to many studies in health care. Linear parametric models, widely used for estimating dose–response relationships, have several limitations. This paper employs the optimal discriminant analysis (ODA) machine‐learning algorithm to determine the degree to which exposure dose can be distinguished based on the distribution of the response variable. By framing the dose–response relationship as a classification problem, machine learning can provide the same functionality as conventional models, but can additionally make individual‐level predictions, which may be helpful in practical applications like establishing responsiveness to prescribed drug regimens. Method: Using data from a study measuring the responses of blood flow in the forearm to the intra‐arterial administration of isoproterenol (separately for 9 black and 13 white men, and pooled), we compare the results estimated from a generalized estimating equations (GEE) model with those estimated using ODA. Results: Generalized estimating equations and ODA both identified many statistically significant dose–response relationships, separately by race and for pooled data. Post hoc comparisons between doses indicated ODA (based on exact P values) was consistently more conservative than GEE (based on estimated P values). Compared with ODA, GEE produced twice as many instances of paradoxical confounding (findings from analysis of pooled data that are inconsistent with findings from analyses stratified by race). Conclusions: Given its unique advantages and greater analytic flexibility, maximum‐accuracy machine‐learning methods like ODA should be considered as the primary analytic approach in dose–response applications. [ABSTRACT FROM AUTHOR]

Background: Pulmonary dysfunction (PDF) in intubated patients remains a serious and costly complication of intensive care unit care. Optimal cardiopulmonary therapy strategies to prevent and manage PDF need clarification to reduce practice variability. The purpose of this paper is to report on the content validation of an evidence-based clinical management algorithm (EBCMA) aimed at the prevention, identification and management of PDF in critically ill patients., Methods: Forty-four draft algorithm statements extracted from the extant literature by the primary research team were verified and rated by research clinicians (n = 7) in an electronic three-round Delphi process. Statements which reached a priori defined consensus [semi-interquartile range (SIQR) <0.5] were collated into the EBCMA., Results: One hundred per cent response rate. Forty-four statements were added after round one. Consensus was reached on rating of 83% (73/88) statements. Differences in interpretation of the existing evidence base, and variations in accepted clinical practice were identified. Four themes were identified where panel failed to reach consensus., Conclusion: The internationally agreed hierarchical framework of current available evidence and clinical expertise developed through this Delphi process provides clinicians with a tool to inform clinical practice. This tool has the potential to reduce practice variability thereby maximizing safety and treatment outcome. The clinical utility of the EBCMA requires further evaluation., (© 2010 Blackwell Publishing Ltd.)

Background Identification of key factors associated with the risk of adverse cardiovascular events and quantification of this risk using multivariable prediction algorithms are among the major advances made in preventive cardiology and cardiovascular epidemiology. Methods In the present paper, we examined clinical predictors of adverse cardiovascular events among 228 individuals with symptoms suggestive of coronary artery disease (CAD) undergoing functional (stress echocardiography) and anatomical (coronary angiography) assessment of CAD. Particularly, we evaluate the possibility to integrate simple measures that have known prognostic value and more recently discovered predictors of risk, such as stress-related ventricular function data and angiographic data, in a unique model implementing a Bayesian network (BN). Moreover, we compared the performance of BN and the covariates hierarchy with those obtained from logistic regression model and from a set of alternative tools becoming popular in various clinical settings, including random forest classification tree analysis, artificial neural networks and support vector machine. Results Network graph and results coming from sensitivity analysis, where variables are ranked according to the gain they provided in variance reduction, seem have an easily intuitive lecture: variables that are measure of ventricular disfunction or of the extent of CAD show a greater impact in predicting event. On the other hand, anamnestic data such as diabetes, dyslipidaemia, hypertension, smoke habits, which are related to the outcome throughout a process of intermediate variables, per se have a small role in outcome prediction. BNs are able to explain a relevant part of variance (70%) and have discrimination ability superior or comparable with those to random forest classification tree analysis, artificial neural networks and support vector machine. Discussion Despite the complexity of interactions, model obtained implementing a BN seems to be able to adequately describe the relationships existing among the analysed variables. BN, being able to predict scenarios in which new variables can be incorporated as health process evolves, can measure individual's risks for adverse cardiovascular events, providing a permanent second opinion to the medical practitioner and assisting diagnostic and therapeutic process. [ABSTRACT FROM AUTHOR]

Rationale: Traditional whole systems of medicine, such as naturopathy, are founded upon holism; a philosophical paradigm consistent with contemporary complexity science. Naturopathic case management is predicated upon the understanding of an intimately interconnected internal physiological and external context of the human organism—potentially indicating a worldview aligned with a complexity perspective. In this study we investigate naturopathic clinical reasoning using a complexity lens with the aim of ascertaining the extent of correspondence between the two. Method: Mind maps depicting case presentations were sought from Australian degree qualified naturopaths. A network mapping was undertaken, which was then analysed in accordance with a complexity science framework using exploratory data analysis and network analysis processes and tools. Results: Naturopathic case schematics, in the form of mind maps (n = 70), were collected, network mapped, and analysed. A total of 739 unique elements and 2724 links were identified across the network. Integral elements across the network were: stress, fatigue, general anxiety, systemic inflammation, gut dysbiosis, and diet. A modularity algorithm detected 11 communities, the primary ones of these representing the nervous system and mood; the gastrointestinal tract, liver, and nutrition; immune function and the immune system; and diet and nutrients. Conclusions: Naturopathic case management is holistic and based on a perspective of an integrated physiology and external context of the human organism. The traditional concept of holism, when subjected to a complexity lens, leads to the emergence of a contemporary holistic paradigm cognisant of the human organism being a complex system. The application of complexity science to investigate naturopathic case management as employed in this study, demonstrates that it is possible to investigate traditional philosophies and principles in a scientific and critical manner. A complexity science research approach may offer a suitable scientific paradigm to develop our understanding of traditional whole systems of medicine. [ABSTRACT FROM AUTHOR]

Rationale, Aims and Objectives: The development of clinical practice guidelines (CPG) suffers from the lack of an explicit and transparent framework for synthesising the key elements necessary to formulate practice recommendations. We matched deliberations of the American Society of Haematology (ASH) CPG panel for the management of pulmonary embolism (PE) with the corresponding decision‐theoretical constructs to assess agreement of the panel recommendations with explicit decision modelling. Methods: Five constructs were identified of which three were used to reformulate the panel's recommendations: (1) standard, expected utility threshold (EUT) decision model; (2) acceptable regret threshold model (ARg) to determine the frequency of tolerable false negative (FN) or false positive (FP) recommendations, and (3) fast‐and‐frugal tree (FFT) decision trees to formulate the entire strategy for management of PE. We compared four management strategies: withhold testing versus d‐dimer → computerized pulmonary angiography (CTPA) ('ASH‐Low') versus CTPA→ d‐dimer ('ASH‐High') versus treat without testing. Results: Different models generated different recommendations. For example, according to EUT, testing should be withheld for prior probability PE < 0.13%, a clinically untenable threshold which is up to 15 times (2/0.13) below the ASH guidelines threshold of ruling out PE (at post probability of PE ≤ 2%). Three models only agreed that the 'ASH low' strategy should be used for the range of pretest probabilities of PE between 0.13% and 13.27% and that the 'ASH high' management should be employed in a narrow range of the prior PE probabilities between 90.85% and 93.07%. For all other prior probabilities of PE, choosing one model did not ensure coherence with other models. Conclusions: CPG panels rely on various decision‐theoretical strategies to develop its recommendations. Decomposing CPG panels' deliberation can provide insights if the panels' deliberation retains a necessary coherence in developing guidelines. CPG recommendations often do not agree with the EUT decision analysis, widely used in medical decision‐making modelling. [ABSTRACT FROM AUTHOR]

Rationale, Aims and Objectives: Critics have charged that evidence‐based medicine (EBM) overemphasises algorithmic rules over unstructured clinical experience and intuition, but the role of structured decision support systems in improving health outcomes remains uncertain. We aim to assess if delivery of anticoagulant prophylaxis in hospitalised patients with COVID‐19 according to an algorithm based on evidence‐based clinical practice guideline (CPG) improved clinical outcomes compared with administration of anticoagulant treatment given at individual practitioners' discretion. Methods: An observational design consisting of the analysis of all acutely ill, consecutive patients (n = 1783) with confirmed COVID‐19 diagnosis admitted between 10 March 2020 to 11 January 2022 to an US academic center. American Society of Haematology CPG for anticoagulant prophylaxis in hospitalised patients with COVID‐19 was converted into a clinical pathway and translated into fast‐and‐frugal decision (FFT) tree ('algorithm'). We compared delivery of anticoagulant prophylaxis in hospitalised patients with COVID‐19 according to the FFT algorithm with administration of anticoagulant treatment given at individual practitioners' discretion. Results: In an adjusted analysis, using combination of Lasso (least absolute shrinkage and selection operator) and propensity score based weighting [augmented inverse‐probability weighting] statistical techniques controlling for cluster data, the algorithm did not reduce death, venous thromboembolism, or major bleeding, but helped avoid longer hospital stay [number of patients needed to be treated (NNT) = 40 (95% CI: 23–143), indicating that for every 40 patients (23–143) managed on FFT algorithm, one avoided staying in hospital longer than 10 days] and averted admission to intensive‐care unit (ICU) [NNT = 19 (95% CI: 13–40)]. All model's selected covariates were well balanced. The results remained robust to sensitivity analyses used to test the stability of the findings. Conclusions: When delivered using a structured FFT algorithm, CPG shortened the hospital stay and help avoided admission to ICU, but it did not affect other relevant outcomes. [ABSTRACT FROM AUTHOR]

Rational and objective: Requests for magnetic resonance imaging (MRI) exams have notably increased in Canada. However, many of these exams may not always be indicated. The Joint Department of Medical Imaging and the eReferral Program have worked collaboratively to embed an integrated clinical decision support (DS) tool within the eReferral process for diagnostic imaging requests. This retrospective chart review aimed to assess the necessity of MRI exams for knee pain patients at the point of referral in relation to the referral method (no DS tools within fax‐ vs. DS tools within eReferral). Methods: Seven hundred and seventeen medical charts of routine MRI referral requests to an Ontario Hospital for patients with knee complaints were reviewed during the study period. The necessity of the MRI exams was evaluated using the supporting algorithm and knee pathway appropriateness guidelines. MRI exams were considered necessary if requested for symptoms or signs that align with best‐practice standards, complemented with sound clinical assessment or history of a radiography scan before ordering an MRI. Results: In general, MRI requests made through eReferral were 13.289 times more likely to be necessary orders than those made through fax. The likelihood of referring patients for a necessary MRI exam was higher for eReferral than fax for the year 2018/2019 (53.0% vs. 26.8%, P < 0.001) and for the year 2019/2020 (58.5% vs. 16.3%, P < 0.001). In addition, the rate of ordering X‐ray as the proper initial imaging scan for patients presenting with knee pain has steadily increased by 10% over the year for users of the eReferral platform compared to a decrease of 7% for those using fax. Conclusion: Our findings highlight the positive impact of integrating DS tools at the point of referral in supporting the ordering of necessary MRI scans, suggesting that service re‐design and implementation of automated assistive technology services would impact patient care. [ABSTRACT FROM AUTHOR]

The article discusses the application and public perception of the use of artificial intelligence (AI) in medicine and public health. Majority of people in the U.S. believe that using AI techniques in decision-making situations is unacceptable and unfair, based on a Pew Research Center report. It explains the factors to consider prior to AI implementation in the health care industry including regulations and ethics. Also cited is the importance of training data monitoring to AI application.

Rationale, aims and objectives: New therapies are increasingly approved by regulatory agencies such as the Food and Drug Administration (FDA) and the European Medicines Agency (EMA) based on testing in non‐randomized clinical trials. These treatments have typically displayed "dramatic effects" (ie, effects that are considered large enough to obviate the combined effects of biases and random errors that may affect the study results). The agencies, however, have not identified how large these effects should be to avoid the need for further testing in randomized controlled trials (RCTs). We investigated the effect size that would circumvent the need for further RCTs testing by the regulatory agencies. We hypothesized that the approval of therapeutic interventions by regulators is based on heuristic decision making whose accuracy can be best characterized by the application of signal detection theory (SDT). Methods: We merged the EMA and FDA database of approvals based on non‐RCT comparisons. We excluded duplicate entries between the two databases. We included a total of 134 approvals of drugs and devices based on non‐RCTs. We integrated Weber‐Fechner law of psychophysics and recognition heuristics within SDT to provide descriptive explanations of the decisions made by the FDA and EMA to approve new treatments based on non‐randomized studies without requiring further testing in RCTs. Results: Our findings suggest that when the difference between novel treatments and the historical control is at least one logarithm (base 10) of magnitude, the veracity of testing in non‐RCTs seems to be established. Conclusion: Drug developers and practitioners alike can use the change in one logarithm of effect size as a benchmark to decide if further testing in RCTs should be pursued, or as a guide to interpreting the results reported in non‐randomized studies. However, further research would be useful to better characterize the threshold of effect size above which testing in RCTs is not needed. [ABSTRACT FROM AUTHOR]

Guidelines, it is assumed in health care circles, have worth. In the ideal health care system they may offer everything from medical certainty to legal protection. They have the potential to save practitioners from the trauma of indecision and at the same time protect them from the consequences of wrong decisions. In this paper, I discuss the impact of guidelines on general practitioners and consumers. Both groups are at the effector end where guidelines should have maximal impact. It is primarily medical practitioners and their constituents in the long run that have to make guidelines work. Theoretical questions about the essential worth of guidelines for consumers and general practitioners are explored, as well as the more practical issues of utility of guidelines. My hypothesis is that guidelines may be conceptually worthwhile, but as yet are of unproven utility. This argument is traced through the literature surrounding the involvement of general practitioners and consumers in guideline development, implementation, review and relevance. From this information a new role for guidelines which is cogniscent of the needs and circumstances of the end users is postulated. Guidelines can become the basis for the principles of sound clinical practice which allow for the unique individual circumstances of clinical practice whilst also providing a consolidated basis for this practice. [ABSTRACT FROM AUTHOR]

Rationale, aims, and objectives: Monash Watch (MW) aims to reduce potentially preventable hospitalisations in a cohort above a risk "threshold" identified by Health Links Chronic Care (HLCC) algorithms using personal, diagnostic, and service data. MW conducted regular patient monitoring through outbound phone calls using the Patient Journey Record System (PaJR). PaJR alerts are intended to act as a self‐reported barometer of stressors, resilience, and health perceptions with more alerts per call indicating greater risk. Aims: To describe predictors of PaJR alerts (self‐reported from outbound phone calls) and predictors of acute admissions based upon a Theoretical Model for Static and Dynamic Indicators of Acute Admissions. Methods: Participants: HLCC cohort with predicted 3+ admissions/year in MW service arm for >40 days; n = 244. Baseline measures—Clinical Frailty Index (CFI); Connor Davis Resilience (CD‐RISC): SF‐12v2 Health Survey scores Mental (MSC) and Physical (PSC) and ICECAP‐O. Dynamic measures: PaJR alerts/call in 10 869 MW records. Acute (non‐surgical) admissions from Victorian Admitted Episode database. Analysis: Logistic regression, correlations, and timeseries homogeneity metrics using XLSTAT. Findings Baseline indicators were significantly correlated except SF‐12_MCS. SF12‐MSC, SF12‐PSC and ICECAP‐O best predicted PaJR alerts/call (ROC: 0.84). CFI best predicted acute admissions (ROC: 0.66), adding CD‐RISC, SF‐12_MCS, SF‐12_PCS and ICECAP‐O with two‐way interactions improved model (ROC: 0.70). PaJR alerts were higher ≤10 days preceding acute admissions and significantly correlated with admissions. Patterns in PaJR alerts in four case studies demonstrated dynamic variations signifying risk. Overall, all baseline indicators were explanatory supporting the theoretical model. Timing of PaJR alerts and acute admissions reflecting changing stressors, resilience, and health perceptions were not predicted from baseline indicators but provided a trigger for service interventions. Conclusion: Both static and dynamic indicators representing stressors, resilience, and health perceptions have the potential to inform threshold models of admission risk in ways that could be clinically useful. [ABSTRACT FROM AUTHOR]

Abstract: Background: Contemporary delivery of health care is inappropriate in many ways, largely due to suboptimal Q5 decision‐making. A typical approach to improve practitioners' decision‐making is to develop evidence‐based clinical practice guidelines (CPG) by guidelines panels, who are instructed to use their judgments to derive practice recommendations. However, mechanisms for the formulation of guideline judgments remains a “black‐box” operation—a process with defined inputs and outputs but without sufficient knowledge of its internal workings. Methods: Increased explicitness and transparency in the process can be achieved by implementing CPG as clinical pathways (CPs) (also known as clinical algorithms or flow‐charts). However, clinical recommendations thus derived are typically ad hoc and developed by experts in a theory‐free environment. As any recommendation can be right (true positive or negative), or wrong (false positive or negative), the lack of theoretical structure precludes the quantitative assessment of the management strategies recommended by CPGs/CPs. Results: To realize the full potential of CPGs/CPs, they need to be placed on more solid theoretical grounds. We believe this potential can be best realized by converting CPGs/CPs within the heuristic theory of decision‐making, often implemented as fast‐and‐frugal (FFT) decision trees. This is possible because FFT heuristic strategy of decision‐making can be linked to signal detection theory, evidence accumulation theory, and a threshold model of decision‐making, which, in turn, allows quantitative analysis of the accuracy of clinical management strategies. Conclusions: Fast‐and‐frugal provides a simple and transparent, yet solid and robust, methodological framework connecting decision science to clinical care, a sorely needed missing link between CPGs/CPs and patient outcomes. We therefore advocate that all guidelines panels express their recommendations as CPs, which in turn should be converted into FFTs to guide clinical care. [ABSTRACT FROM AUTHOR]

Abstract: Rationale, aims and objectives: A common approach to assessing treatment effects in nonrandomized studies with time‐to‐event outcomes is to estimate propensity scores and compute weights using logistic regression, test for covariate balance, and then estimate treatment effects using Cox regression. A machine‐learning alternative—classification tree analysis (CTA)—used to generate propensity scores and to estimate treatment effects in time‐to‐event data may identify complex relationships between covariates not found using conventional regression‐based approaches. Method: Using empirical data, we identify all statistically valid CTA propensity score models and then use them to compute strata‐specific, observation‐level propensity score weights that are subsequently applied in outcomes analyses. We compare findings obtained using this framework to the conventional method, by evaluating covariate balance and treatment effect estimates obtained using Cox regression and a weighted CTA outcomes model. Results: All models had some imbalanced covariates. Nevertheless, treatment effect estimates were generally consistent across all weighted models. Conclusions: In the study sample, given that all approaches elicited similar results, using CTA increased confidence that bias could not be reduced any further. Because the CTA algorithm identifies all statistically valid propensity score models for a sample, it is most likely to identify a correctly specified propensity score model—and therefore should be used either to confirm results using traditional methods, or to reveal biases that may be missed by traditional methods. Moreover, given that the true treatment effect is never known in observational data, CTA should be considered for estimating outcomes because no statistical assumptions are required. [ABSTRACT FROM AUTHOR]

Abstract: Rationale, aims and objectives: Clinicians considering improvements in health‐related quality of life (HRQoL) after total hip replacement (THR) must account for multiple pieces of information. Evidence‐based decisions are important to best assess the effect of THR on HRQoL. This work aims at constructing a shared decision‐making tool that helps clinicians assessing the future benefits of THR by offering predictions of 1‐year postoperative HRQoL of THR patients. Methods: We used data from the Swedish Hip Arthroplasty Register. Data from 2008 were used as training set and data from 2009 to 2012 as validation set. We adopted two approaches. First, we assumed a continuous distribution for the EQ‐5D index and modelled the postoperative EQ‐5D index with regression models. Second, we modelled the five dimensions of the EQ‐5D and weighted together the predictions using the UK Time Trade‐Off value set. As predictors, we used preoperative EQ‐5D dimensions and the EQ‐5D index, EQ visual analogue scale, visual analogue scale pain, Charnley classification, age, gender, body mass index, American Society of Anesthesiologists, surgical approach and prosthesis type. Additionally, the tested algorithms were combined in a single predictive tool by stacking. Results: Best predictive power was obtained by the multivariate adaptive regression splines (R2 = 0.158). However, this was not significantly better than the predictive power of linear regressions (R2 = 0.157). The stacked model had a predictive power of 17%. Conclusions: Successful implementation of a shared decision‐making tool that can aid clinicians and patients in understanding expected improvement in HRQoL following THR would require higher predictive power than we achieved. For a shared decision‐making tool to succeed, further variables, such as socioeconomics, need to be considered. [ABSTRACT FROM AUTHOR]

Rationale, aims and objectives In evaluating non-randomized interventions, propensity scores (PS) estimate the probability of assignment to the treatment group given observed characteristics. Machine learning algorithms have been proposed as an alternative to conventional logistic regression for modelling PS in order to avoid limitations of linear methods. We introduce classification tree analysis (CTA) to generate PS which is a 'decision-tree'-like classification model that provides accurate, parsimonious decision rules that are easy to display and interpret, reports P values derived via permutation tests, and evaluates cross-generalizability. Method Using empirical data, we identify all statistically valid CTA PS models and then use them to compute strata-specific, observation-level PS weights that are subsequently applied in outcomes analyses. We compare findings obtained using this framework to logistic regression and boosted regression, by evaluating covariate balance using standardized differences, model predictive accuracy, and treatment effect estimates obtained using median regression and a weighted CTA outcomes model. Results While all models had some imbalanced covariates, main-effects logistic regression yielded the lowest average standardized difference, whereas CTA yielded the greatest predictive accuracy. Nevertheless, treatment effect estimates were generally consistent across all models. Conclusions Assessing standardized differences in means as a test of covariate balance is inappropriate for machine learning algorithms that segment the sample into two or more strata. Because the CTA algorithm identifies all statistically valid PS models for a sample, it is most likely to identify a correctly specified PS model, and should be considered as an alternative approach to modeling the PS. [ABSTRACT FROM AUTHOR]

Rationale, aims, and objectives Stratification is a popular propensity score (PS) adjustment technique. It has been shown that stratifying the PS into 5 quantiles can remove over 90% of the bias due to the covariates used to generate the PS. Because of this finding, many investigators partition their data into 5 quantiles of the PS without examining whether a more robust solution (one that increases covariate balance while potentially reducing bias in the outcome analysis) can be found for their data. Two approaches (referred to herein as PSCORE and PSTRATA) obtain the optimal stratification solution by repeatedly dividing the data into strata until balance is achieved between treatment and control groups on the PS. These algorithms differ in how they partition the data, and it is not known which is better, or if either is better than a 5-quantile default approach, for reducing bias in treatment effect estimates. Method Monte Carlo simulations and empirical data are used to assess whether PS strata defined by PSCORE, PSTRATA, or 5 quantiles is best at reducing bias in treatment effect estimates, when used within a marginal mean weighting framework (MMWS). These estimates are further compared to results derived using inverse probability of treatment weights (IPTW). Results PSTRATA was slightly better than PSCORE in balancing covariates and reducing bias, while both approaches outperformed the 5-quantile approach. Overall MMWS using any stratification method outperformed IPTW. Conclusions Investigators should routinely use stratification approaches that obtain the optimal stratification solution, rather than simply partitioning the data into 5 quantiles of the PS. Moreover, MMWS (in conjunction with an optimal stratification approach) should be considered as an alternative to IPTW in studies that use PS weights. [ABSTRACT FROM AUTHOR]

Rationale, aims and objectives Hospital readmission within 30 days of discharge occurs in almost 20% of US Medicare patients and may be a marker of poor quality inpatient care, ineffective hospital to home transitions, or disease severity. Within a patient centered medical home, care transition interventions may only be practical from cost and staffing perspectives if targeted at patients with the greatest risk of readmission. Various scoring algorithms attempt to predict patients at risk for 30-day readmission, but head-to-head comparison of performance is lacking. Compare published scoring algorithms which use generally available electronic medical record data on the same set of hospitalized primary care patients. Methods The LACE index, the LACE+ index, the HOSPITAL score, and the readmission risk score were computed on a consecutive cohort of 26,278 hospital admissions. Classifier performance was assessed by plotting receiver operating characteristic curves comparing the computed score with the actual outcome of death or readmission within 30 days. Statistical significance of differences in performance was assessed using bootstrapping techniques. Results Correct readmission classification on this cohort was moderate with the following c-statistics: Readmission risk score 0.666; LACE 0.680; LACE+ 0.662; and HOSPITAL 0.675. There was no statistically significant difference in performance between classifiers. Conclusions Logistic regression based classifiers yield only moderate performance when utilized to predict 30-day readmissions. The task is difficult due to the variety of underlying causes for readmission, nonlinearity, and the arbitrary time period of concern. More sophisticated classification techniques may be necessary to increase performance and allow patient centered medical homes to effectively focus efforts to reduce readmissions. [ABSTRACT FROM AUTHOR]

Rationale, aims and objectives: Reducing the rate of hospital readmissions, particularly avoidable ones, has significant implications on patient outcomes, cost containment, and quality of care. Given that the reason of readmission may differ from the patient's main diagnosis in the index admission, this study aims to assess the influence of index comorbidities on the primary readmission diagnoses and explore the risk of deemed avoidable readmission because of prior comorbidities. Methods: A retrospective review of 3962 discharges was conducted at a 527-bed teaching hospital in Jordan, utilizing data related to 2025 internal medicine patients. Results: Among all discharges, 29% were followed by a 30-day readmission, of which 13% were identified as potentially avoidable. Of all readmissions, 36% of patients were readmitted because of one of the comorbidities that had been identified at index admission. In addition, 47% of the potentially avoidable readmissions had a main diagnosis that was one of the index comorbidities. The results also showed an association between readmission for one of the index stay's comorbidities and being avoidable, with an adjusted odds ratio of 2.12 (95% confidence interval, 1.65-2.72). Overall, the presence of certain diseases, being identified as one of the preceding comorbidities, was found to have a substantial influence on the risk of potentially avoidable readmission. These diseases included digestive, circulatory, respiratory, genitourinary systems, and infectious and parasitic diseases (adjusted relative risks = 1.57,1.49, 1.36,1.30, and 2.30, respectively). Conclusion: To help reduce the rates of readmission, potential gains seem available if hospitals adopt clinical practices that support the patient's care during the post-discharge transition. This implies that health care providers need to pay more attention to the comorbidities of high-risk patients to be closely monitored after discharge. [ABSTRACT FROM AUTHOR]

A response by Halfdan Petursson and colleagues to a letter to the editor about their article "Is the use of cholesterol in mortality risk algorithms in clinical guidelines valid? Ten years prospective data from the Norwegian HUNT 2 study" is presented.

Rationale, aims and objectives The focus on the diagnosis is a pivotal aspect of medical practice since antiquity. Diagnostic taxonomy helped to categorize ailments to improve medical care, and in its social sense resulted in validation of the sick role for some, but marginalization or stigmatization for others. In the medical industrial complex, diagnostic taxonomy structured health care financing, management and practitioner remuneration. However, with increasing demands from multiple agencies, there are increasing unintended and unwarranted consequences of our current taxonomies and diagnostic processes resulting from the conglomeration of underpinning concepts, theories, information and motivations. Results We argue that the increasing focus on the diagnosis resulted in excessive compartmentalization - 'partialism' - of medical practice, diminishing medical care and being naively simplistic in light of the emerging understanding of the interconnected nature of the diseasome. The human is a complex organic system of interconnecting dynamics and feedback loops responding to internal and external forces including genetic, epigenetic and environmental attractors, rather than the sum of multiple discrete organs which can develop isolated diseases or multiple morbidities. Solutions to these unintended consequences of many contemporary health system processes involve revisiting the nature of diagnostic taxonomies and the processes of their construction. A dynamic taxonomic framework would shift to more relevant attractors at personal, clinical and health system levels recognizing the non-linear nature of health and disease. Human health at an individual, group and population level is the ability to adapt to internal and external stressors with resilience throughout the life course, yet diagnostic taxonomies are increasingly constructed around fixed anchors. Conclusions Understanding diagnosis as dissecting, pigeonholing or bean counting ( learning by dividing) is no longer useful, the challenge for the future is to understand the big picture ( learning by connecting). Diagnostic categorization needs to embrace a meta-learning approach open to human variability. [ABSTRACT FROM AUTHOR]

Rationale, aims and objectives The goal is to fully automate the calculation of a breast ptosis measure from clinical photographs through automatic localization of fiducial points relevant to the measure. Methods Sixty-eight women (97 clinical photographs) who underwent or were scheduled for breast reconstruction were included. The photographs were divided into a development set ( N = 49) and an evaluation set ( N = 48). The breast ptosis measure is obtained automatically from distances between three fiducial points: the nipple, the lowest visible point of breast ( LVP), and the lateral terminus of the inframammary fold ( LT). The nipple is localized using the YIQ colour space to highlight the contrast between the areola and the surrounding breast skin. The areola is localized using its shape, location and high Q component intensity. The breast contour is estimated using Dijkstra's shortest path algorithm on the gradient of the photograph in greyscale. The lowest point of the estimated contour is set as the LVP. To locate the anatomically subtle LT, the location of patient's axilla is used as a reference. Results The algorithm's efficacy was evaluated by comparing manual and automated localizations of the fiducial points. The average nipple diameter was used as a cut-off to define success. The algorithm showed 90, 91 and 83% accuracy for locating the nipple, LVP and LT in the evaluation set, respectively. Conclusion This study presents a new automated algorithm that may facilitate the quantification of breast ptosis from lateral views of patients' photographs. [ABSTRACT FROM AUTHOR]

Rationale, aims and objectives This study aimed to determine the value of using a mix of clinical pharmacy data and routine hospital admission spell data in the development of predictive algorithms. Exploration of risk factors in hospitalized patients, together with the targeting strategies devised, will enable the prioritization of clinical pharmacy services to optimize patient outcomes. Methods Predictive algorithms were developed using a number of detailed steps using a 75% sample of integrated medicines management ( IMM) patients, and validated using the remaining 25%. IMM patients receive targeted clinical pharmacy input throughout their hospital stay. The algorithms were applied to the validation sample, and predicted risk probability was generated for each patient from the coefficients. Risk threshold for the algorithms were determined by identifying the cut-off points of risk scores at which the algorithm would have the highest discriminative performance. Clinical pharmacy staffing levels were obtained from the pharmacy department staffing database. Results Numbers of previous emergency admissions and admission medicines together with age-adjusted co-morbidity and diuretic receipt formed a 12-month post-discharge and/or readmission risk algorithm. Age-adjusted co-morbidity proved to be the best index to predict mortality. Increased numbers of clinical pharmacy staff at ward level was correlated with a reduction in risk-adjusted mortality index ( RAMI). Conclusions Algorithms created were valid in predicting risk of in-hospital and post-discharge mortality and risk of hospital readmission 3, 6 and 12 months post-discharge. The provision of ward-based clinical pharmacy services is a key component to reducing RAMI and enabling the full benefits of pharmacy input to patient care to be realized. [ABSTRACT FROM AUTHOR]

Objectives This study demonstrates the development, reliability and outcome of a targeted method for standardized assessment of adverse drug events ( ADEs) in surgical patients. Methods Initial practice evaluation of this ADE assessment method was carried out in a prospective single centre cohort study. In total, 262 electively admitted surgical patients were included. The surgical trigger tool was applied to patients' medical records by two independent reviewers, and subsequent assessment of causality, severity and preventability of ADEs was carried out by two independent expert panels consisting of a consultant surgeon and a clinical pharmacologist. The surgical trigger tool and causality assessment method were each tested on reliability in a separate group of 50 randomly selected patients using Fleiss and Cohen's kappa statistics and percentages of agreement. Comparison of this method with an existing trigger tool method for ADEs was performed. Results Our surgical trigger tool contains 51 triggers. The inter- and intra-rater calculations showed substantial to almost perfect levels of agreement (kappa range 0.71-0.83), with a 97.8-98.5% percentage of agreement. Fair to substantial levels of agreement were calculated for causality, severity and preventability (kappa range 0.38-0.79). The percentages of inter- and intra-rater agreement were 68.9 and 70.5% for causality, 67.0 and 82.0% for severity, and both 98.4% for preventability, respectively. Compared with the existing trigger tool method for ADEs, we found an additional 363 triggers, 18 ADEs (an extra 20%) and 3 preventable ADEs in our surgical cohort. Conclusions This targeted trigger tool for standardized assessment of ADEs in surgical patients shows excellent agreement between reviewers. The assessment of medication-related harm had acceptable agreement. Compared with an existing ADE trigger tool method, the present method found almost 20% extra ADEs. This method can be a useful alternative to existing trigger tool methods, in particular to assess medication safety in surgical patients. [ABSTRACT FROM AUTHOR]

Objective To test the feasibility of classifying emergency department patients into severity grades using data mining methods. Design Emergency department records of 402 patients were classified into five severity grades by two expert physicians. The Naïve Bayes and C4.5 algorithms were applied to produce classifiers from patient data into severity grades. The classifiers' results over several subsets of the data were compared with the physicians' assessments, with a random classifier, and with a classifier that selects the maximal-prevalence class. Measurements Positive predictive value, multiple-class extensions of sensitivity and specificity combinations, and entropy change. Results The mean accuracy of the data mining classifiers was 52.94 ± 5.89%, significantly better ( P < 0.05) than the mean accuracy of a random classifier (34.60 ± 2.40%). The entropy of the input data sets was reduced through classification by a mean of 10.1%. Allowing for classification deviations of one severity grade led to mean accuracy of 85.42 ± 1.42%. The classifiers' accuracy in that case was similar to the physicians' consensus rate. Learning from consensus records led to better performance. Reducing the number of severity grades improved results in certain cases. The performance of the Naïve Bayes and C4.5 algorithms was similar; in unbalanced data sets, Naïve Bayes performed better. Conclusions It is possible to produce a computerized classification model for the severity grade of triage patients, using data mining methods. Learning from patient records regarding which there is a consensus of several physicians is preferable to learning from each physician's patients. Either Naïve Bayes or C4.5 can be used; Naïve Bayes is preferable for unbalanced data sets. An ambiguity in the intermediate severity grades seems to hamper both the physicians' agreement and the classifiers' accuracy. [ABSTRACT FROM AUTHOR]

Rationale, aims and objectives The quality of life of breast cancer survivors is ameliorated by minimizing adverse effects on their physical appearance. Breast reconstruction is important for restoring the survivor's appearance. In breast reconstructive surgery, there is a need to develop technologies for quantifying surgical outcomes and understanding women's perceptions of changes in their appearance. Methods for objectively measuring breast anatomy are needed in order to help breast cancer survivors, radiation oncologists and surgeons quantify changes in appearance that occur with different breast reconstructive surgical options. Methods In this study, we present an automated method for computing a variant of the normalized Breast Retraction Assessment, a common measure of symmetry, from routine clinical photographs taken to document breast cancer treatment procedures. The algorithms were designed using a development set of retrospectively collected clinical photographs and evaluated using a separate test set of prospectively collected clinical photographs. Results We have developed new algorithms for automatically localizing the umbilicus and nipples on standard anterior-posterior clinical photographs, enabling the automatic calculation of measure of breast symmetry. Conclusions This study demonstrates that under standardized photographic conditions, automatic localization of fiducial points and subsequent computation of measures, such as symmetry, is feasible. The algorithms presented here for quantifying surgical outcomes in breast reconstructive surgery will provide a foundation for future work on assisting a cancer patient and her surgeons in selecting and planning reconstruction procedures that will maximize the woman's psychosocial adjustment to life as a breast cancer survivor. [ABSTRACT FROM AUTHOR]

Abstract Medicine has evolved toward rationalization since the Enlightenment, favouring quantitative measures. Now, a paradigm shift toward control through formalization can be observed in health care whose structures and processes are subjected to increasing standardization. However, educational reforms and curricula do not yet adequately respond to this shift. The aim of this article is to describe innovative approaches in medical education for adapting to these changes. The study design is a descriptive case report relying on a literature review and on a reform project’s evaluation. Concept mapping is used to graphically represent relationships among concepts, i.e. defined terms from educational literature. Definitions of ‘concept map’, ‘guideline’ and ‘algorithm’ are presented. A prototypical algorithm for organizational decision making in the project’s instructional design is shown. Evaluation results of intrinsic learning motivation are demonstrated: intrinsic learning motivation depends upon students’ perception of their competence exhibiting path coefficients varying from 0.42 to 0.51. Perception of competence varies with the type of learning environment. An innovative educational format, called ‘evidence-based learning (EBL)’ is deduced from these findings and described here. Effects of formalization consist of structuring decision making about implementation of different learning environments or about minimizing variance in teaching or learning. Unintended effects of formalization such as implementation problems and bureaucracy are discussed. Formalized tools for designing medical education are available. Specific instructional designs influence students’ learning motivation. Concept maps are suitable for controlling educational quality, thus enabling the paradigm shift in medical education. [ABSTRACT FROM AUTHOR]