Heuristic decision making in medicine

Development and Launch of a Comprehensive Fish and Wildlife Reporting Mobile Application

This paper describes the conception, development, launch, marketing, and maintenance of the Florida Fish & Wildlife Conservation Commission Reporter (FWC Reporter), a mobile application (app) that the public can use to report fish and wildlife and environmental concerns to appropriate state agency offices. The FWC Reporter, launched in February 2018, is the first comprehensive reporting app developed by a fish and wildlife agency in the United States that we are aware of. On the basis of a consumer concept test and library research, the app's design incorporates elements of fast and frugal heuristics and inductive decision theory. The FWC developed this version of the app in-house with an open-source development package, and involves primarily email and telephone to send reports and facilitate communication with agency staff. From February 2018 through December 31, 2018, the FWC recorded over 1,000 downloads on Android and 4,308 on Apple devices, and 258 emailed reports. Since fish and wildlife reporting apps do not provide clear benefit to the user (low value proposition), it is crucial that offices receiving reports are proactive in building and maintaining their own reporting constituency. Engagement with the public is not only through active recruitment of app users, but also following up after a report is made so that the users know their efforts are appreciated. Two FWC programs demonstrated successful engagement, horseshoe crabs Limulus polyphemus and fish kills, as they received the most email reports. Other topics for any conservation or fish and wildlife agency to consider when developing their own reporting mobile app include download intention, adoption, continuance, habit formation, public and organizational acceptance, and marketing.

Machine-learning-derived rules set excludes risk of Parkinson's disease in patients with olfactory or gustatory symptoms with high accuracy

BACKGROUND

Chemosensory loss is a symptom of Parkinson's disease starting already at preclinical stages. Their appearance without an identifiable etiology therefore indicates a possible early symptom of Parkinson's disease. Supervised machine-learning was used to identify parameters that predict Parkinson's disease among patients having sought medical advice for chemosensory symptoms.

METHODS

Olfactory, gustatory and demographic parameters were analyzed in 247 patients who had reported for chemosensory symptoms. Unsupervised machine-learning, implanted as so-called fast and frugal decision trees, was applied to map these parameters to a diagnosis of Parkinson's disease queried for in median 9 years after the first interview.

RESULTS

A symbolic hierarchical decision rule-based classifier was created that comprised d = 5 parameters, including scores in tests of odor discrimination, odor identification and olfactory thresholds, the age at which the chemosensory loss has been noticed, and a familial history of Parkinson's disease. The rule set provided a cross-validated negative predictive performance of Parkinson's disease of 94.1%; however, its balanced accuracy to predict the disease was only 58.9% while robustly above guessing.

CONCLUSIONS

Applying machine-learning techniques, a classifier was developed that took the shape of a set of six hierarchical rules with binary decisions about olfaction-related features or a familial burden of Parkinson's disease. Its main clinical strength lies in the exclusion of the possibility of developing Parkinson's disease in a patient with olfactory or gustatory loss.

An Electronic-Based Curriculum to Train Acute Care Providers in Rural Haiti and India

BACKGROUND

Access to a trained, competent health care workforce remains a challenge globally, particularly in rural settings. To bridge this gap, the World Health Organization calls for innovations in electronic learning and task shifting. Yet, these approaches are underutilized due to cost, challenges associated with implementing technology, and a lack of suitably educated trainees.

OBJECTIVE

We explored the feasibility of the Acute Care Providers Project (ACPP) to remotely train community members to be health care providers in 2 sites: Haiti and India.

METHODS

The ACP program is an asynchronous curriculum that provides core health content and a structured approach to clinical care through an electronic curriculum. The curriculum is reinforced with case-based practice and hands-on workshops for procedural skills. ACPP was deployed in rural Haiti and India. Evaluation of the program included multiple-choice pretests and posttests, an objective structured clinical examination (OSCE), and direct observation of skills.

RESULTS

Four Haitian and 55 Indian trainees completed the course. In Haiti, mean scores were 34.8% (SD 12.4) on the pretest and 78.0% (SD 6.5) on the posttest (P = .004). Trainees scored 100% on the OSCE and passed the skills checklist. In India, mean scores were 16.5% (SD 3.9) on the pretest and 81.7% (SD 9.0) on the posttest (P < .001). Trainees scored a median of 91.8% (SD 3.95) on the OSCE and all passed the skills checklist.

CONCLUSIONS

The ACPP offers a scalable, replicable asynchronous curriculum to train lay individuals to provide basic health care in rural communities.

Determining qualitative effect size ratings using a likelihood ratio scatter matrix in diagnostic test accuracy systematic reviews

Background Diagnostic test accuracy (DTA) systematic reviews (SRs) characterize a test's potential for diagnostic quality and safety. However, interpreting DTA measures in the context of SRs is challenging. Further, some evidence grading methods (e.g. Centers for Disease Control and Prevention, Division of Laboratory Systems Laboratory Medicine Best Practices method) require determination of qualitative effect size ratings as a contributor to practice recommendations. This paper describes a recently developed effect size rating approach for assessing a DTA evidence base. Methods A likelihood ratio scatter matrix will plot positive and negative likelihood ratio pairings for DTA studies. Pairings are graphed as single point estimates with confidence intervals, positioned in one of four quadrants derived from established thresholds for test clinical validity. These quadrants support defensible judgments on "substantial", "moderate", or "minimal" effect size ratings for each plotted study. The approach is flexible in relation to a priori determinations of the relative clinical importance of false positive and false negative test results. Results and conclusions This qualitative effect size rating approach was operationalized in a recent SR that assessed effectiveness of test practices for the diagnosis of Clostridium difficile. Relevance of this approach to other methods of grading evidence, and efforts to measure diagnostic quality and safety are described. Limitations of the approach arise from understanding that a diagnostic test is not an isolated element in the diagnostic process, but provides information in clinical context towards diagnostic quality and safety.

Decision making with visualizations: a cognitive framework across disciplines

Visualizations-visual representations of information, depicted in graphics-are studied by researchers in numerous ways, ranging from the study of the basic principles of creating visualizations, to the cognitive processes underlying their use, as well as how visualizations communicate complex information (such as in medical risk or spatial patterns). However, findings from different domains are rarely shared across domains though there may be domain-general principles underlying visualizations and their use. The limited cross-domain communication may be due to a lack of a unifying cognitive framework. This review aims to address this gap by proposing an integrative model that is grounded in models of visualization comprehension and a dual-process account of decision making. We review empirical studies of decision making with static two-dimensional visualizations motivated by a wide range of research goals and find significant direct and indirect support for a dual-process account of decision making with visualizations. Consistent with a dual-process model, the first type of visualization decision mechanism produces fast, easy, and computationally light decisions with visualizations. The second facilitates slower, more contemplative, and effortful decisions with visualizations. We illustrate the utility of a dual-process account of decision making with visualizations using four cross-domain findings that may constitute universal visualization principles. Further, we offer guidance for future research, including novel areas of exploration and practical recommendations for visualization designers based on cognitive theory and empirical findings.

Guiding practitioners through end of life care for people with dementia: The use of heuristics

Background

End of life care (EOLC) for people with dementia can present a multitude of challenges and difficult decisions for practitioners. These challenges may include assessment and management of difficulties with eating and swallowing, responding to agitation, treating pain, and managing recurrent infections. Practitioners sometimes lack both confidence in making end of life decisions and guidance. This study developed an alternative to lengthy guidelines, in the form of heuristics which were tested in clinical settings. The aim of this study was to test the usability and acceptability of a set of heuristics which could be used by practitioners providing EOLC for people with dementia in a variety of clinical and care settings.

Methods

A three phase co-design process was adopted: 1) Synthesis of evidence and outputs from interviews and focus groups with family carers and practitioners, by a co-design group, to develop heuristics; 2) Testing of the heuristics in five clinical or care settings for six months; 3) Evaluation of the heuristics at three and six months using qualitative individual and group interviews.

Results

Four heuristics were developed covering: eating and swallowing difficulties, agitation and restlessness, reviewing treatment and interventions at the end of life, and providing routine care. The five sites reported that the heuristics were simple and easy to use, comprehensive, and made implicit, tacit knowledge explicit. Four themes emerged from the qualitative evaluation: authority and permission; synthesis of best practice; providing a structure and breaking down complexity; and reassurance and instilling confidence.

Conclusion

Use of heuristics is a novel approach to end of life decision making in dementia which can be useful to both experienced and junior members of staff making decisions. Heuristics are a practical tool which could overcome a lack of care pathways and direct guidance in end of life care for people with dementia.

Deferral rate variability in blood donor eligibility assessment

BACKGROUND

Both donors and the blood bank rely on the result of the donor health interview. However, survey data suggest that substantial variability in deferral rates among interviewers exist. We studied whether variability remained after adjusting for conditional factors.

STUDY DESIGN AND METHODS

The data set included Dutch interview data on whole blood donor visits in 2015, where one of their visits was selected randomly. We applied logistic regression and multilevel regression analyses with the donor visit, with the interviewer representing the levels. We set up four models: 1) all reasons deferral, 2) low‐hemoglobin‐level deferral, 3) infectious disease risk deferral and 4) other medical reasons deferral.

RESULTS

In total, 138,398 visits were included in the study, of which 60,534 (43.7%) related to male donors. The overall deferral rate for men was 7.91% and for women 12.25%. Deferral rates among interviewers ranged from as low as 1.19% up to 28.8%. Models 2 (low hemoglobin level) and particularly 4 (other medical reasons), for both men and women, showed significant intraclass correlation coefficients, implying considerable deferral rate variability among interviewers. Donor age, the number of previous visits, and the season had relatively large effects. However, explained variances of the logistic regression models were relatively low, ranging from 2.53% to 7.35%.

CONCLUSION

Deferral appears to be a random process, while substantial variability was found among interviewer deferral rates, suggesting that some interviewers are more cautious than others. Our results suggest heuristic and subjective diagnosing to be prevalent. Steps should be taken to improve interview result validity.

Perception of domestic risks among carers for dependent older persons

BACKGROUND

Despite growing interest for home care, little evidence exists on the perception of domestic risk among carers for dependent older persons. This study aims to characterize the risks to which carers of aging dependent individuals are exposed, and to determine whether these risk dimensions are predictive for effective support, for burden, and for psychological distress.

METHODS

Seventy care partners were questioned about the risk situations identified at the homes of the old people they care for, about the burden they felt in their role, and about their feelings of psychological distress. Securing was evaluated by means of sensibility measures, and overprotection was evaluated by means of specificity measures.

RESULTS

Risk rates were high for loneliness of the old people, wandering, burns, and unsatisfactory health monitoring. There was very little overlap between identification of the risks and implementation of solutions by the caregiver, except for the risks that involved heat. The distinction between accurate securing and overprotection is especially important, because the burden of care partners was linked to uncontrolled domestic risks.

CONCLUSION

Typologies of reactions to risk, characterized by a signal detection approach, could contribute to a better understanding of the situations experienced by care partners, especially situations of neglect and of overprotection.

External validity and anchoring heuristics: application of DUNDRUM-1 to secure service gatekeeping in South Wales

Aims and method Structured clinical judgement tools provide scope for the standardisation of forensic service gatekeeping and also allow identification of heuristics in this decision process. The DUNDRUM-1 triage tool was completed retrospectively for 121 first-time referrals to forensic services in South Wales. Fifty were admitted to medium security, 49 to low security and 22 remained in open conditions.

RESULTS

DUNDRUM-1 total scores differed appropriately between different levels of security. However, regression revealed heuristic anchoring on the 'legal process' and 'immediacy of risk due to mental disorder' items. Clinical implications Patient placement was broadly aligned with DUNDRUM-1 recommendations. However, not all triage items informed gatekeeping decisions. It remains to be seen whether decisions anchored in this way are effective. Declaration of interest Dr Mark Freestone gave permission for AUC values from Freestone et al. (2015) to be presented here for comparison.

PATIENT-CENTERED DECISION MAKING: LESSONS FROM MULTI-CRITERIA DECISION ANALYSIS FOR QUANTIFYING PATIENT PREFERENCES

OBJECTIVES

Patient preferences should be a central consideration in healthcare decision making. However, stories of patients challenging regulatory and reimbursement decisions has led to questions on whether patient voices are being considered sufficiently during those decision making processes. This has led some to argue that it is necessary to quantify patient preferences before they can be adequately considered.

METHODS

This study considers the lessons from the use of multi-criteria decision analysis (MCDA) for efforts to quantify patient preferences. It defines MCDA and summarizes the benefits it can provide to decision makers, identifies examples of MCDAs that have involved patients, and summarizes good practice guidelines as they relate to quantifying patient preferences.

RESULTS

The guidance developed to support the use of MCDA in healthcare provide some useful considerations for the quantification of patient preferences, namely that researchers should give appropriate consideration to: the heterogeneity of patient preferences, and its relevance to decision makers; the cognitive challenges posed by different elicitation methods; and validity of the results they produce. Furthermore, it is important to consider how the relevance of these considerations varies with the decision being supported.

CONCLUSIONS

The MCDA literature holds important lessons for how patient preferences should be quantified to support healthcare decision making.

Hospital-Based Physicians' Intubation Decisions and Associated Mental Models when Managing a Critically and Terminally Ill Older Patient

BACKGROUND

Variation in the intensity of acute care treatment at the end of life is influenced more strongly by hospital and provider characteristics than patient preferences.

OBJECTIVE

We sought to describe physicians' mental models (i.e., thought processes) when encountering a simulated critically and terminally ill older patient, and to compare those models based on whether their treatment plan was patient preference-concordant or preference-discordant.

METHODS

Seventy-three hospital-based physicians from 3 academic medical centers engaged in a simulated patient encounter and completed a mental model interview while watching the video recording of their encounter. We used an "expert" model to code the interviews. We then used Kruskal-Wallis tests to compare the weighted mental model themes of physicians who provided preference-concordant treatment with those who provided preference-discordant treatment.

RESULTS

Sixty-six (90%) physicians provided preference-concordant treatment and 7 (10%) provided preference-discordant treatment (i.e., they intubated the patient). Physicians who intubated the patient were more likely to emphasize the reversible and emergent nature of the patient situation (z = -2.111, P = 0.035), their own comfort (z = -2.764, P = 0.006), and rarely focused on explicit patient preferences (z = 2.380, P = 0.017).

LIMITATIONS

Post-decisional interviewing with audio/video prompting may induce hindsight bias. The expert model has not yet been validated and may not be exhaustive. The small sample size limits generalizability and power.

CONCLUSIONS

Hospital-based physicians providing preference-discordant used a different mental model for decision making for a critically and terminally ill simulated case. These differences may offer targets for future interventions to promote preference-concordant care for seriously ill patients.

Developing a patient-reported outcome measure for HIV care on perceived barriers to antiretroviral adherence: assessing the needs of HIV clinicians through qualitative analysis

PURPOSE

To identify HIV clinicians' needs for the clinical use of a new patient-reported outcome measure (PRO) on barriers to antiretroviral therapy (ART) adherence.

METHODS

In 2015, five focus groups with 31 clinicians from France were transcribed, coded with Atlas.ti, and submitted to a typological analysis.

RESULTS

The analysis identified seven patient profiles, each tied to distinct barriers to adherence and to specific needs for the PRO's content, data collection and transmission. Clinicians preferred, for the patient who is: (1) 'passive,' that the PRO collect information on ART knowledge, to ensure that the prescription's instructions are being respected; (2) 'misleading,' that it be able to detect adherence to ART and socially desirable responses; (3) 'stoic,' that questions challenge the patient to recognize treatment-specific side effects; (4) 'hedonistic,' that the PRO contains content on lifestyle and risk-taking; (5) 'obsessive,' that the PRO captures quality of life and stressful life events; (6) 'overburdened,' that the PRO provides information on the person's home environment, socioeconomic status and cultural constraints. For all or most patient profiles, the clinicians wished that the PRO be completed, minimally, prior to the medical consultation and to receive alerts, under varying conditions, when problematic scores were detected. Depending on the profile, there was preference for the inclusion of open-ended questions and transmission of cross-sectional, periodic or longitudinal PRO data.

CONCLUSION

Overall, this study's findings suggest that to support the clinical management of ART adherence, our PRO must meet the needs of a wide variety of patients and must perform multiple functions.

Analyzing How Discursive Practices Affect Physicians' Decision-Making Processes: A Phenomenological-Based Qualitative Study in Critical Care Contexts

An intensive care unit (ICU) is a demanding environment, defined by significant complexity, in which physicians must make decisions in situations characterized by high levels of uncertainty. This study used a phenomenological approach to investigate the decision-making (DM) processes among ICU physicians’ team with the aim of understanding what happens when ICU physicians must reach a decision about the infectious status of a patient. The focus was put on the identification of how the discursive practices influence physicians’ DM processes and on how different ICU environments make different discursive profiles emerge, particularly when a key issue is at the center of the physicians’ discussion. A naturalistic approach used in this study is particularly suitable for investigating health care practices because it can best illuminate the essential meaning of the “lived experiences” of the participants. The findings revealed a common framework of elements that provide insight into DM processes in ICUs and how these are affected by discursive practices.

A qualitative study exploring factors influencing clinical decision-making for influenza-like illness in Solapur city, Maharashtra, India

The co-existence of different types of medical systems (medical pluralism) is a typical feature of India's healthcare system. For conditions such as influenza-like illness (ILI), where non-specific disease signs/symptoms exist, clinical reasoning in the context of medical pluralism becomes crucial. Recognising this need, we undertook a qualitative study, which explored factors underpinning clinical decisions on diagnosis and management of ILI. The study involved semi-structured interviews including clinical vignettes with 20 healthcare practitioners (working within allopathy, homeopathy and Ayurveda) working in the private healthcare sector in Solapur city, India. An inquiry was conducted into criteria influencing the diagnosis, treatment, referral to specialist care and role of treatment guidelines for ILI. Thematic analysis was used to identify aspects relating to ILI diagnosis, treatment and referral. The diagnosis of influenza was based largely on clinical symptoms suggestive of influenza in the absence of other diagnoses. Referral for laboratory tests was only initiated if illness did not resolve, generally after 2-3 consultations. Antibiotics were often prescribed for persistent illness, with antivirals rarely considered. Some differences between practitioners from different medical systems were observed in relation to treatment and referral in case of persistent illness. A combination of analytical and intuitive clinical reasoning was used by the participants and clinical decisions were based on both social and clinical factors. Clinical decision-making was rarely a linear process and respondents felt that broad guidelines on influenza that allowed doctors to account for the sociocultural context within which they practised medicine would be helpful.

Introduction to MAchine Learning & Knowledge Extraction (MAKE)

The grand goal of Machine Learning is to develop software which can learn from previous experience—similar to how we humans do. Ultimately, to reach a level of usable intelligence, we need (1) to learn from prior data, (2) to extract knowledge, (3) to generalize—i.e., guessing where probability function mass/density concentrates, (4) to fight the curse of dimensionality, and (5) to disentangle underlying explanatory factors of the data—i.e., to make sense of the data in the context of an application domain. To address these challenges and to ensure successful machine learning applications in various domains an integrated machine learning approach is important. This requires a concerted international effort without boundaries, supporting collaborative, cross-domain, interdisciplinary and transdisciplinary work of experts from seven sections, ranging from data pre-processing to data visualization, i.e., to map results found in arbitrarily high dimensional spaces into the lower dimensions to make it accessible, usable and useful to the end user. An integrated machine learning approach needs also to consider issues of privacy, data protection, safety, security, user acceptance and social implications. This paper is the inaugural introduction to the new journal of MAchine Learning & Knowledge Extraction (MAKE). The goal is to provide an incomplete, personally biased, but consistent introduction into the concepts of MAKE and a brief overview of some selected topics to stimulate future research in the international research community.

Navigating the Decision Space: Shared Medical Decision Making as Distributed Cognition

Despite increasing prominence, little is known about the cognitive processes underlying shared decision making. To investigate these processes, we conceptualize shared decision making as a form of distributed cognition. We introduce a Decision Space Model to identify physical and social influences on decision making. Using field observations and interviews, we demonstrate that patients and physicians in both acute and chronic care consider these influences when identifying the need for a decision, searching for decision parameters, making actionable decisions Based on the distribution of access to information and actions, we then identify four related patterns: physician dominated; physician-defined, patient-made; patient-defined, physician-made; and patient-dominated decisions. Results suggests that (a) decision making is necessarily distributed between physicians and patients, (b) differential access to information and action over time requires participants to transform a distributed task into a shared decision, and (c) adverse outcomes may result from failures to integrate physician and patient reasoning. Our analysis unifies disparate findings in the medical decision-making literature and has implications for improving care and medical training.

[Cognitive errors in diagnostic decision making]

Approximately 10-15% of our diagnostic decisions are faulty and may lead to unfavorable and dangerous outcomes, which could be avoided. These diagnostic errors are mainly caused by cognitive biases in the diagnostic reasoning process.Our medical diagnostic decision-making is based on intuitive "System 1" and analytical "System 2" diagnostic decision-making and can be deviated by unconscious cognitive biases.These deviations can be positively influenced on a systemic and an individual level. For the individual, metacognition (internal withdrawal from the decision-making process) and debiasing strategies, such as verification, falsification and rule out worst-case scenarios, can lead to improved diagnostic decisions making.

Artificial Neural Network for Total Laboratory Automation to Improve the Management of Sample Dilution

Diluting a sample to obtain a measure within the analytical range is a common task in clinical laboratories. However, for urgent samples, it can cause delays in test reporting, which can put patients' safety at risk. The aim of this work is to show a simple artificial neural network that can be used to make it unnecessary to predilute a sample using the information available through the laboratory information system. Particularly, the Multilayer Perceptron neural network built on a data set of 16,106 cardiac troponin I test records produced a correct inference rate of 100% for samples not requiring predilution and 86.2% for those requiring predilution. With respect to the inference reliability, the most relevant inputs were the presence of a cardiac event or surgery and the result of the previous assay. Therefore, such an artificial neural network can be easily implemented into a total automation framework to sensibly reduce the turnaround time of critical orders delayed by the operation required to retrieve, dilute, and retest the sample.

Constructing diagnostic likelihood: clinical decisions using subjective versus statistical probability

BACKGROUND

Although physicians are highly trained in the application of evidence-based medicine, and are assumed to make rational decisions, there is evidence that their decision making is prone to biases. One of the biases that has been shown to affect accuracy of judgements is that of representativeness and base-rate neglect, where the saliency of a person's features leads to overestimation of their likelihood of belonging to a group. This results in the substitution of 'subjective' probability for statistical probability.

METHODS

This study examines clinicians' propensity to make estimations of subjective probability when presented with clinical information that is considered typical of a medical condition. The strength of the representativeness bias is tested by presenting choices in textual and graphic form. Understanding of statistical probability is also tested by omitting all clinical information.

RESULTS

For the questions that included clinical information, 46.7% and 45.5% of clinicians made judgements of statistical probability, respectively. Where the question omitted clinical information, 79.9% of clinicians made a judgement consistent with statistical probability. There was a statistically significant difference in responses to the questions with and without representativeness information (χ2 (1, n=254)=54.45, p<0.0001).

CONCLUSIONS

Physicians are strongly influenced by a representativeness bias, leading to base-rate neglect, even though they understand the application of statistical probability. One of the causes for this representativeness bias may be the way clinical medicine is taught where stereotypic presentations are emphasised in diagnostic decision making.

Big data for bipolar disorder

The delivery of psychiatric care is changing with a new emphasis on integrated care, preventative measures, population health, and the biological basis of disease. Fundamental to this transformation are big data and advances in the ability to analyze these data. The impact of big data on the routine treatment of bipolar disorder today and in the near future is discussed, with examples that relate to health policy, the discovery of new associations, and the study of rare events. The primary sources of big data today are electronic medical records (EMR), claims, and registry data from providers and payers. In the near future, data created by patients from active monitoring, passive monitoring of Internet and smartphone activities, and from sensors may be integrated with the EMR. Diverse data sources from outside of medicine, such as government financial data, will be linked for research. Over the long term, genetic and imaging data will be integrated with the EMR, and there will be more emphasis on predictive models. Many technical challenges remain when analyzing big data that relates to size, heterogeneity, complexity, and unstructured text data in the EMR. Human judgement and subject matter expertise are critical parts of big data analysis, and the active participation of psychiatrists is needed throughout the analytical process.

Cognitive biases associated with medical decisions: a systematic review

Background

Cognitive biases and personality traits (aversion to risk or ambiguity) may lead to diagnostic inaccuracies and medical errors resulting in mismanagement or inadequate utilization of resources. We conducted a systematic review with four objectives: 1) to identify the most common cognitive biases, 2) to evaluate the influence of cognitive biases on diagnostic accuracy or management errors, 3) to determine their impact on patient outcomes, and 4) to identify literature gaps.

Methods

We searched MEDLINE and the Cochrane Library databases for relevant articles on cognitive biases from 1980 to May 2015. We included studies conducted in physicians that evaluated at least one cognitive factor using case-vignettes or real scenarios and reported an associated outcome written in English. Data quality was assessed by the Newcastle-Ottawa scale. Among 114 publications, 20 studies comprising 6810 physicians met the inclusion criteria. Nineteen cognitive biases were identified.

Results

All studies found at least one cognitive bias or personality trait to affect physicians. Overconfidence, lower tolerance to risk, the anchoring effect, and information and availability biases were associated with diagnostic inaccuracies in 36.5 to 77 % of case-scenarios. Five out of seven (71.4 %) studies showed an association between cognitive biases and therapeutic or management errors. Of two (10 %) studies evaluating the impact of cognitive biases or personality traits on patient outcomes, only one showed that higher tolerance to ambiguity was associated with increased medical complications (9.7 % vs 6.5 %; p = .004). Most studies (60 %) targeted cognitive biases in diagnostic tasks, fewer focused on treatment or management (35 %) and on prognosis (10 %). Literature gaps include potentially relevant biases (e.g. aggregate bias, feedback sanction, hindsight bias) not investigated in the included studies. Moreover, only five (25 %) studies used clinical guidelines as the framework to determine diagnostic or treatment errors. Most studies (n = 12, 60 %) were classified as low quality.

Conclusions

Overconfidence, the anchoring effect, information and availability bias, and tolerance to risk may be associated with diagnostic inaccuracies or suboptimal management. More comprehensive studies are needed to determine the prevalence of cognitive biases and personality traits and their potential impact on physicians’ decisions, medical errors, and patient outcomes.

Electronic supplementary material

The online version of this article (doi:10.1186/s12911-016-0377-1) contains supplementary material, which is available to authorized users.

Predicting the Transition From Acute Stress Disorder to Posttraumatic Stress Disorder in Children With Severe Injuries

INTRODUCTION

The purpose of this study was to examine predictors of risk for and the transition between acute stress disorder (ASD) and posttraumatic stress disorder (PTSD) in a longitudinal sample of youth with severe injuries admitted to the hospital. These data would assist with treatment and discharge planning.

METHODS

Youth were assessed for ASD during the initial hospital stay and were followed-up over an 18-month period for PTSD (n = 151). Youth were classified into four groups, including Resilient (ASD-, PTSD-), ASD Only (ASD+, PTSD-), PTSD Only (ASD-, PTSD+), and Chronic (ASD+, PTSD+). Demographic, psychiatric, social context, and injury-related factors were examined as predictors of diagnostic transition.

RESULTS

The results of multivariate analysis of variance and pairwise comparisons found that peritraumatic dissociation, gender, and socioeconomic status were significant predictors after controlling for multiple testing.

DISCUSSION

Results suggest that both within-child and contextual factors contribute to the longitudinal response to trauma in children. Clinicians should consider early screening and discharge planning, particularly for children most at risk.

A co-design process developing heuristics for practitioners providing end of life care for people with dementia

BACKGROUND

The end of life for someone with dementia can present many challenges for practitioners; such as, providing care if there are swallowing difficulties. This study aimed to develop a toolkit of heuristics (rules-of-thumb) to aid practitioners making end-of-life care decisions for people with dementia.

METHODS

An iterative co-design approach was adopted using a literature review and qualitative methods, including; 1) qualitative interviews and focus groups with family carers and 2) focus groups with health and care professionals. Family carers were recruited from a national charity, purposively sampling those with experience of end-of-life care for a person with dementia. Health and care professionals were purposively sampled to include a broad range of expertise including; general practitioners, palliative care specialists, and geriatricians. A co-design group was established consisting of health and social care experts and family carers, to synthesise the findings from the qualitative work and produce a toolkit of heuristics to be tested in practice.

RESULTS

Four broad areas were identified as requiring complex decisions at the end of life; 1) eating/swallowing difficulties, 2) agitation/restlessness, 3) ending life-sustaining treatment, and 4) providing "routine care" at the end of life. Each topic became a heuristic consisting of rules arranged into flowcharts. Eating/swallowing difficulties have three rules; ensuring eating/swallowing difficulties do not come as a surprise, considering if the situation is an emergency, and considering 'comfort feeding' only versus time-trialled artificial feeding. Agitation/restlessness encourages a holistic approach, considering the environment, physical causes, and the carer's wellbeing. Ending life-sustaining treatment supports practitioners through a process of considering the benefits of treatment versus quality-of-life and comfort. Finally, a heuristic on providing routine care such as bathing, prompts practitioners to consider adapting the delivery of care, in order to promote comfort and dignity at the end of life.

CONCLUSIONS

The heuristics are easy to use and remember, offering a novel approach to decision making for dementia end-of-life care. They have the potential to be used alongside existing end-of-life care recommendations, adding more readily available practical assistance. This is the first study to synthesise experience and existing evidence into easy-to-use heuristics for dementia end-of-life care.

Choosing Wisely® in Preventive Medicine: The American College of Preventive Medicine's Top 5 List of Recommendations

The Choosing Wisely(®) initiative is a national campaign led by the American Board of Internal Medicine Foundation, focused on quality improvement and advancing a dialogue on avoiding wasteful or unnecessary medical tests, procedures, and treatments. The American College of Preventive Medicine (ACPM) Prevention Practice Committee is an active participant in the Choosing Wisely project. The committee created the ACPM Choosing Wisely Task Force to lead the development of ACPM's recommendations with the intention of facilitating wise decisions about the appropriate use of preventive care. After utilizing an iterative process that involved reviewing evidence-based literature, the ACPM Choosing Wisely Task Force developed five recommendations targeted toward overused services within the field of preventive medicine. These include: (1) don't take a multivitamin, vitamin E, or beta carotene to prevent cardiovascular disease or cancer; (2) don't routinely perform prostate-specific antigen-based screening for prostate cancer; (3) don't use whole-body scans for early tumor detection in asymptomatic patients; (4) don't use expensive medications when an equally effective and lower-cost medication is available; and (5) don't perform screening for cervical cancer in low-risk women aged 65 years or older and in women who have had a total hysterectomy for benign disease. The Task Force also reviewed some of the barriers to implementing these recommendations, taking into account the interplay between system and environmental characteristics, and identified specific strategies necessary for timely utilization of these recommendations.

"First, know thyself": cognition and error in medicine

Although error is an integral part of the world of medicine, physicians have always been little inclined to take into account their own mistakes and the extraordinary technological progress observed in the last decades does not seem to have resulted in a significant reduction in the percentage of diagnostic errors. The failure in the reduction in diagnostic errors, notwithstanding the considerable investment in human and economic resources, has paved the way to new strategies which were made available by the development of cognitive psychology, the branch of psychology that aims at understanding the mechanisms of human reasoning. This new approach led us to realize that we are not fully rational agents able to take decisions on the basis of logical and probabilistically appropriate evaluations. In us, two different and mostly independent modes of reasoning coexist: a fast or non-analytical reasoning, which tends to be largely automatic and fast-reactive, and a slow or analytical reasoning, which permits to give rationally founded answers. One of the features of the fast mode of reasoning is the employment of standardized rules, termed "heuristics." Heuristics lead physicians to correct choices in a large percentage of cases. Unfortunately, cases exist wherein the heuristic triggered fails to fit the target problem, so that the fast mode of reasoning can lead us to unreflectively perform actions exposing us and others to variable degrees of risk. Cognitive errors arise as a result of these cases. Our review illustrates how cognitive errors can cause diagnostic problems in clinical practice.

The Role of Big Data in the Management of Sleep-Disordered Breathing

Analysis of large-volume data holds promise for improving the application of precision medicine to sleep, including improving identification of patient subgroups who may benefit from alternative therapies. Big data used within the health care system also promises to facilitate end-to-end screening, diagnosis, and management of sleep disorders; improve the recognition of differences in presentation and susceptibility to sleep apnea; and lead to improved management and outcomes. To meet the vision of personalized, precision therapeutics and diagnostics and improving the efficiency and quality of sleep medicine will require ongoing efforts, investments, and change in our current medical and research cultures.

A calibration hierarchy for risk models was defined: from utopia to empirical data

OBJECTIVE

Calibrated risk models are vital for valid decision support. We define four levels of calibration and describe implications for model development and external validation of predictions.

STUDY DESIGN AND SETTING

We present results based on simulated data sets.

RESULTS

A common definition of calibration is "having an event rate of R% among patients with a predicted risk of R%," which we refer to as "moderate calibration." Weaker forms of calibration only require the average predicted risk (mean calibration) or the average prediction effects (weak calibration) to be correct. "Strong calibration" requires that the event rate equals the predicted risk for every covariate pattern. This implies that the model is fully correct for the validation setting. We argue that this is unrealistic: the model type may be incorrect, the linear predictor is only asymptotically unbiased, and all nonlinear and interaction effects should be correctly modeled. In addition, we prove that moderate calibration guarantees nonharmful decision making. Finally, results indicate that a flexible assessment of calibration in small validation data sets is problematic.

CONCLUSION

Strong calibration is desirable for individualized decision support but unrealistic and counter productive by stimulating the development of overly complex models. Model development and external validation should focus on moderate calibration.

What Do Physicians Believe About the Way Decisions Are Made? A Pilot Study on Metacognitive Knowledge in the Medical Context

Metacognition relative to medical decision making has been poorly investigated to date. However, beliefs about methods of decision making (metacognition) play a fundamental role in determining the efficiency of the decision itself. In the present study, we investigated a set of beliefs that physicians develop in relation to the modes of making decisions in a professional environment. The Solomon Questionnaire, designed to assess metacognitive knowledge about behaviors and mental processes involved in decision making, was administered to a sample of 18 emergency physicians, 18 surgeons, and 18 internists. Significant differences in metacognitive knowledge emerged among these three medical areas. Physicians' self-reports about the decision process mirrored the peculiarities of the context in which they operate. Their metacognitive knowledge demonstrated a reflective attitude that is an effective tool during the decision making process.

What Do Physicians Believe About the Way Decisions Are Made? A Pilot Study on Metacognitive Knowledge in the Medical Context

Metacognition relative to medical decision making has been poorly investigated to date. However, beliefs about methods of decision making (metacognition) play a fundamental role in determining the efficiency of the decision itself. In the present study, we investigated a set of beliefs that physicians develop in relation to the modes of making decisions in a professional environment. The Solomon Questionnaire, designed to assess metacognitive knowledge about behaviors and mental processes involved in decision making, was administered to a sample of 18 emergency physicians, 18 surgeons, and 18 internists. Significant differences in metacognitive knowledge emerged among these three medical areas. Physicians' self-reports about the decision process mirrored the peculiarities of the context in which they operate. Their metacognitive knowledge demonstrated a reflective attitude that is an effective tool during the decision making process.

After the Liverpool Care Pathway--development of heuristics to guide end of life care for people with dementia: protocol of the ALCP study

INTRODUCTION

End of life care guidance for people with dementia is lacking and this has been made more problematic in England with the removal of one of the main end of life care guidelines which offered some structure, the Liverpool Care Pathway. This guidance gap may be eased with the development of heuristics (rules of thumb) which offer a fast and frugal form of decision-making.

OBJECTIVE

To develop a toolkit of heuristics (rules of thumb) for practitioners to use when caring for people with dementia at the end of life.

METHOD AND ANALYSIS

A mixed-method study using a co-design approach to develop heuristics in three phases. In phase 1, we will conduct at least six focus groups with family carers, health and social care practitioners from both hospital and community care services, using the 'think-aloud' method to understand decision-making processes and to develop a set of heuristics. The focus group topic guide will be developed from the findings of a previous study of 46 interviews of family carers about quality end-of-life care for people with dementia and a review of the literature. A multidisciplinary development team of health and social care practitioners will synthesise the findings from the focus groups to devise and refine a toolkit of heuristics. Phase 2 will test the use of heuristics in practice in five sites: one general practice, one community nursing team, one hospital ward and two palliative care teams working in the community. Phase 3 will evaluate and further refine the toolkit of heuristics through group interviews, online questionnaires and semistructured interviews.

ETHICS AND DISSEMINATION

This study has received ethical approval from a local NHS research ethics committee (Rec ref: 15/LO/0156). The findings of this study will be presented in peer-reviewed publications and national and international conferences.

Exploring the Potential of Predictive Analytics and Big Data in Emergency Care

Clinical research often focuses on resource-intensive causal inference, whereas the potential of predictive analytics with constantly increasing big data sources remains largely unexplored. Basic prediction, divorced from causal inference, is much easier with big data. Emergency care may benefit from this simpler application of big data. Historically, predictive analytics have played an important role in emergency care as simple heuristics for risk stratification. These tools generally follow a standard approach: parsimonious criteria, easy computability, and independent validation with distinct populations. Simplicity in a prediction tool is valuable, but technological advances make it no longer a necessity. Emergency care could benefit from clinical predictions built using data science tools with abundant potential input variables available in electronic medical records. Patients' risks could be stratified more precisely with large pools of data and lower resource requirements for comparing each clinical encounter to those that came before it, benefiting clinical decisionmaking and health systems operations. The largest value of predictive analytics comes early in the clinical encounter, in which diagnostic and prognostic uncertainty are high and resource-committing decisions need to be made. We propose an agenda for widening the application of predictive analytics in emergency care. Throughout, we express cautious optimism because there are myriad challenges related to database infrastructure, practitioner uptake, and patient acceptance. The quality of routinely compiled clinical data will remain an important limitation. Complementing big data sources with prospective data may be necessary if predictive analytics are to achieve their full potential to improve care quality in the emergency department.

Do We Know Why We Make Errors in Morphological Diagnosis? An Analysis of Approach and Decision-Making in Haematological Morphology

BACKGROUND

The laboratory interpretation of blood film morphology is frequently a rapid, accurate, and cost-effective final-stage of blood count analysis. However, the interpretation of findings often rests with a single individual, and errors can carry significant impact. Cell identification and classification skills are well supported by existing resources, but the contribution and importance of other skills are less well understood.

METHODS

The UK external quality assurance group in haematology (UK NEQAS(H)) runs a Continued Professional Development scheme where large digital-images of abnormal blood smears are presented using a web-based virtual microscope. Each case is answered by more than 800 individuals. Morphological feature selection and prioritisation, as well as diagnosis and proposed action, are recorded. We analysed the responses of participants, aiming to identify successful strategies as well as sources of error.

FINDINGS

The approach to assessment by participants depended on the affected cell type, case complexity or skills of the morphologist. For cases with few morphological abnormalities, we found that accurate cell identification and classification were the principle requirements for success. For more complex films however, feature recognition and prioritisation had primary importance. Additionally however, we found that participants employed a range of heuristic techniques to support their assessment, leading to associated bias and error.

INTERPRETATION

A wide range of skills together allow successful morphological assessment and the complexity of this process is not always understood or recognised. Heuristic techniques are widely employed to support or reinforce primary observations and to simplify complex findings. These approaches are effective and are integral to assessment; however they may also be a source of bias or error. Improving outcomes and supporting diagnosis require the development of decision-support mechanisms that identify and support the benefits of heuristic strategies while identifying or avoiding associated biases.

FUNDING

The CPD scheme is funded by participant subscription.

Heuristics: foundations for a novel approach to medical decision making

Medical decision-making is a complex process that often takes place during uncertainty, that is, when knowledge, time, and resources are limited. How can we ensure good decisions? We present research on heuristics-simple rules of thumb-and discuss how medical decision-making can benefit from these tools. We challenge the common view that heuristics are only second-best solutions by showing that they can be more accurate, faster, and easier to apply in comparison to more complex strategies. Using the example of fast-and-frugal decision trees, we illustrate how heuristics can be studied and implemented in the medical context. Finally, we suggest how a heuristic-friendly culture supports the study and application of heuristics as complementary strategies to existing decision rules.

Risk attitudes and personality traits predict perceptions of benefits and risks for medicinal products: a field study of European medical assessors

BACKGROUND

Risk attitudes and personality traits are known predictors of decision making among laypersons, but very little is known of their influence among experts participating in organizational decision making.

METHODS

Seventy-five European medical assessors were assessed in a field study using the Domain Specific Risk Taking scale and the Big Five Inventory scale. Assessors rated the risks and benefits for a mock "clinical dossier" specific to their area of expertise, and ordinal regression models were used to assess the odds of risk attitude or personality traits in predicting either the benefit or the risk ratings.

RESULTS

An increase in the "conscientiousness" score predicted an increase in the perception of the drug's benefit, and male assessors gave higher scores for the drug's benefit ratings than did female assessors. Extraverted assessors saw fewer risks, and assessors with a perceived neutral-averse or averse risk profile saw greater risks.

CONCLUSIONS

Medical assessors perceive the benefits and risks of medicines via a complex interplay of the medical situation, their personality traits and even their gender. Further research in this area is needed to determine how these potential biases are managed within the regulatory setting.