Safety of patient-facing digital symptom checkers

Evaluation of AI-Driven LabTest Checker for Diagnostic Accuracy and Safety: Prospective Cohort Study

Background

In recent years, the implementation of artificial intelligence (AI) in health care is progressively transforming medical fields, with the use of clinical decision support systems (CDSSs) as a notable application. Laboratory tests are vital for accurate diagnoses, but their increasing reliance presents challenges. The need for effective strategies for managing laboratory test interpretation is evident from the millions of monthly searches on test results' significance. As the potential role of CDSSs in laboratory diagnostics gains significance, however, more research is needed to explore this area.

Objective

The primary objective of our study was to assess the accuracy and safety of LabTest Checker (LTC), a CDSS designed to support medical diagnoses by analyzing both laboratory test results and patients' medical histories.

Methods

This cohort study embraced a prospective data collection approach. A total of 101 patients aged ≥18 years, in stable condition, and requiring comprehensive diagnosis were enrolled. A panel of blood laboratory tests was conducted for each participant. Participants used LTC for test result interpretation. The accuracy and safety of the tool were assessed by comparing AI-generated suggestions to experienced doctor (consultant) recommendations, which are considered the gold standard.

Results

The system achieved a 74.3% accuracy and 100% sensitivity for emergency safety and 92.3% sensitivity for urgent cases. It potentially reduced unnecessary medical visits by 41.6% (42/101) and achieved an 82.9% accuracy in identifying underlying pathologies.

Conclusions

This study underscores the transformative potential of AI-based CDSSs in laboratory diagnostics, contributing to enhanced patient care, efficient health care systems, and improved medical outcomes. LTC's performance evaluation highlights the advancements in AI's role in laboratory medicine.

Evaluating the diagnostic and triage performance of digital and online symptom checkers for the presentation of myocardial infarction; A retrospective cross-sectional study

Online symptom checkers are increasingly popular health technologies that enable patients to input their symptoms to produce diagnoses and triage advice. However, there is concern regarding the performance and safety of symptom checkers in diagnosing and triaging patients with life-threatening conditions. This retrospective cross-sectional study aimed to evaluate and compare commercially available symptom checkers for performance in diagnosing and triaging myocardial infarctions (MI). Symptoms and biodata of MI patients were inputted into 8 symptom checkers identified through a systematic search. Anonymised clinical data of 100 consecutive MI patients were collected from a tertiary coronary intervention centre between 1st January 2020 to 31st December 2020. Outcomes included (1) diagnostic sensitivity as defined by symptom checkers outputting MI as the primary diagnosis (D1), or one of the top three (D3), or top five diagnoses (D5); and (2) triage sensitivity as defined by symptom checkers outputting urgent treatment recommendations. Overall D1 sensitivity was 48±31% and varied between symptom checkers (range: 6-85%). Overall D3 and D5 sensitivity were 73±20% (34-92%) and 79±14% (63-94%), respectively. Overall triage sensitivity was 83±13% (55-91%). 24±16% of atypical cases had a correct D1 though for female atypical cases D1 sensitivity was only 10%. Atypical MI D3 and D5 sensitivity were 44±21% and 48±24% respectively and were significantly lower than typical MI cases (p<0.01). Atypical MI triage sensitivity was significantly lower than typical cases (53±20% versus 84±15%, p<0.01). Female atypical cases had significantly lower diagnostic and triage sensitivity than typical female MI cases (p<0.01).Given the severity of the pathology, the diagnostic performance of symptom checkers for correctly diagnosing an MI is concerningly low. Moreover, there is considerable inter-symptom checker performance variation. Patients presenting with atypical symptoms were under-diagnosed and under-triaged, especially if female. This study highlights the need for improved clinical performance, equity and transparency associated with these technologies.

Does an App a Day Keep the Doctor Away? AI Symptom Checker Applications, Entrenched Bias, and Professional Responsibility

The growing prominence of artificial intelligence (AI) in mobile health (mHealth) has given rise to a distinct subset of apps that provide users with diagnostic information using their inputted health status and symptom information-AI-powered symptom checker apps (AISympCheck). While these apps may potentially increase access to health care, they raise consequential ethical and legal questions. This paper will highlight notable concerns with AI usage in the health care system, further entrenchment of preexisting biases in the health care system and issues with professional accountability. To provide an in-depth analysis of the issues of bias and complications of professional obligations and liability, we focus on 2 mHealth apps as examples-Babylon and Ada. We selected these 2 apps as they were both widely distributed during the COVID-19 pandemic and make prominent claims about their use of AI for the purpose of assessing user symptoms. First, bias entrenchment often originates from the data used to train AI systems, causing the AI to replicate these inequalities through a "garbage in, garbage out" phenomenon. Users of these apps are also unlikely to be demographically representative of the larger population, leading to distorted results. Second, professional accountability poses a substantial challenge given the vast diversity and lack of regulation surrounding the reliability of AISympCheck apps. It is unclear whether these apps should be subject to safety reviews, who is responsible for app-mediated misdiagnosis, and whether these apps ought to be recommended by physicians. With the rapidly increasing number of apps, there remains little guidance available for health professionals. Professional bodies and advocacy organizations have a particularly important role to play in addressing these ethical and legal gaps. Implementing technical safeguards within these apps could mitigate bias, AIs could be trained with primarily neutral data, and apps could be subject to a system of regulation to allow users to make informed decisions. In our view, it is critical that these legal concerns are considered throughout the design and implementation of these potentially disruptive technologies. Entrenched bias and professional responsibility, while operating in different ways, are ultimately exacerbated by the unregulated nature of mHealth.

Controlling Inputter Variability in Vignette Studies Assessing Web-Based Symptom Checkers: Evaluation of Current Practice and Recommendations for Isolated Accuracy Metrics

BACKGROUND

The rapid growth of web-based symptom checkers (SCs) is not matched by advances in quality assurance. Currently, there are no widely accepted criteria assessing SCs' performance. Vignette studies are widely used to evaluate SCs, measuring the accuracy of outcome. Accuracy behaves as a composite metric as it is affected by a number of individual SC- and tester-dependent factors. In contrast to clinical studies, vignette studies have a small number of testers. Hence, measuring accuracy alone in vignette studies may not provide a reliable assessment of performance due to tester variability.

OBJECTIVE

This study aims to investigate the impact of tester variability on the accuracy of outcome of SCs, using clinical vignettes. It further aims to investigate the feasibility of measuring isolated aspects of performance.

METHODS

Healthily's SC was assessed using 114 vignettes by 3 groups of 3 testers who processed vignettes with different instructions: free interpretation of vignettes (free testers), specified chief complaints (partially free testers), and specified chief complaints with strict instruction for answering additional symptoms (restricted testers). κ statistics were calculated to assess agreement of top outcome condition and recommended triage. Crude and adjusted accuracy was measured against a gold standard. Adjusted accuracy was calculated using only results of consultations identical to the vignette, following a review and selection process. A feasibility study for assessing symptom comprehension of SCs was performed using different variations of 51 chief complaints across 3 SCs.

RESULTS

Intertester agreement of most likely condition and triage was, respectively, 0.49 and 0.51 for the free tester group, 0.66 and 0.66 for the partially free group, and 0.72 and 0.71 for the restricted group. For the restricted group, accuracy ranged from 43.9% to 57% for individual testers, averaging 50.6% (SD 5.35%). Adjusted accuracy was 56.1%. Assessing symptom comprehension was feasible for all 3 SCs. Comprehension scores ranged from 52.9% and 68%.

CONCLUSIONS

We demonstrated that by improving standardization of the vignette testing process, there is a significant improvement in the agreement of outcome between testers. However, significant variability remained due to uncontrollable tester-dependent factors, reflected by varying outcome accuracy. Tester-dependent factors, combined with a small number of testers, limit the reliability and generalizability of outcome accuracy when used as a composite measure in vignette studies. Measuring and reporting different aspects of SC performance in isolation provides a more reliable assessment of SC performance. We developed an adjusted accuracy measure using a review and selection process to assess data algorithm quality. In addition, we demonstrated that symptom comprehension with different input methods can be feasibly compared. Future studies reporting accuracy need to apply vignette testing standardization and isolated metrics.

Evaluating the Diagnostic Performance of Symptom Checkers: Clinical Vignette Study

BACKGROUND

Medical self-diagnostic tools (or symptom checkers) are becoming an integral part of digital health and our daily lives, whereby patients are increasingly using them to identify the underlying causes of their symptoms. As such, it is essential to rigorously investigate and comprehensively report the diagnostic performance of symptom checkers using standard clinical and scientific approaches.

OBJECTIVE

This study aims to evaluate and report the accuracies of a few known and new symptom checkers using a standard and transparent methodology, which allows the scientific community to cross-validate and reproduce the reported results, a step much needed in health informatics.

METHODS

We propose a 4-stage experimentation methodology that capitalizes on the standard clinical vignette approach to evaluate 6 symptom checkers. To this end, we developed and peer-reviewed 400 vignettes, each approved by at least 5 out of 7 independent and experienced primary care physicians. To establish a frame of reference and interpret the results of symptom checkers accordingly, we further compared the best-performing symptom checker against 3 primary care physicians with an average experience of 16.6 (SD 9.42) years. To measure accuracy, we used 7 standard metrics, including M1 as a measure of a symptom checker's or a physician's ability to return a vignette's main diagnosis at the top of their differential list, F1-score as a trade-off measure between recall and precision, and Normalized Discounted Cumulative Gain (NDCG) as a measure of a differential list's ranking quality, among others.

RESULTS

The diagnostic accuracies of the 6 tested symptom checkers vary significantly. For instance, the differences in the M1, F1-score, and NDCG results between the best-performing and worst-performing symptom checkers or ranges were 65.3%, 39.2%, and 74.2%, respectively. The same was observed among the participating human physicians, whereby the M1, F1-score, and NDCG ranges were 22.8%, 15.3%, and 21.3%, respectively. When compared against each other, physicians outperformed the best-performing symptom checker by an average of 1.2% using F1-score, whereas the best-performing symptom checker outperformed physicians by averages of 10.2% and 25.1% using M1 and NDCG, respectively.

CONCLUSIONS

The performance variation between symptom checkers is substantial, suggesting that symptom checkers cannot be treated as a single entity. On a different note, the best-performing symptom checker was an artificial intelligence (AI)-based one, shedding light on the promise of AI in improving the diagnostic capabilities of symptom checkers, especially as AI keeps advancing exponentially.

"That's just Future Medicine" - a qualitative study on users' experiences of symptom checker apps

BACKGROUND

Symptom checker apps (SCAs) are mobile or online applications for lay people that usually have two main functions: symptom analysis and recommendations. SCAs ask users questions about their symptoms via a chatbot, give a list with possible causes, and provide a recommendation, such as seeing a physician. However, it is unclear whether the actual performance of a SCA corresponds to the users' experiences. This qualitative study investigates the subjective perspectives of SCA users to close the empirical gap identified in the literature and answers the following main research question: How do individuals (healthy users and patients) experience the usage of SCA, including their attitudes, expectations, motivations, and concerns regarding their SCA use?

METHODS

A qualitative interview study was chosen to clarify the relatively unknown experience of SCA use. Semi-structured qualitative interviews with SCA users were carried out by two researchers in tandem via video call. Qualitative content analysis was selected as methodology for the data analysis.

RESULTS

Fifteen interviews with SCA users were conducted and seven main categories identified: (1) Attitudes towards findings and recommendations, (2) Communication, (3) Contact with physicians, (4) Expectations (prior to use), (5) Motivations, (6) Risks, and (7) SCA-use for others.

CONCLUSIONS

The aspects identified in the analysis emphasise the specific perspective of SCA users and, at the same time, the immense scope of different experiences. Moreover, the study reveals ethical issues, such as relational aspects, that are often overlooked in debates on mHealth. Both empirical and ethical research is more needed, as the awareness of the subjective experience of those affected is an essential component in the responsible development and implementation of health apps such as SCA.

TRIAL REGISTRATION

German Clinical Trials Register (DRKS): DRKS00022465. 07/08/2020.

[Short paths to diagnosis with artificial intelligence: systematic literature review on diagnostic decision support systems]

BACKGROUND

Rare diseases are often recognized late. Their diagnosis is particularly challenging due to the diversity, complexity and heterogeneity of clinical symptoms. Computer-aided diagnostic aids, often referred to as diagnostic decision support systems (DDSS), are promising tools for shortening the time to diagnosis. Despite initial positive evaluations, DDSS are not yet widely used, partly due to a lack of integration with existing clinical or practice information systems.

OBJECTIVE

This article provides an insight into currently existing diagnostic support systems that function without access to electronic patient records and only require information that is easily obtainable.

MATERIALS AND METHODS

A systematic literature search identified eight articles on DDSS that can assist in the diagnosis of rare diseases with no need for access to electronic patient records or other information systems in practices and hospitals. The main advantages and disadvantages of the identified rare disease diagnostic support systems were extracted and summarized.

RESULTS

Symptom checkers and DDSS based on portrait photos and pain drawings already exist. The degree of maturity of these applications varies.

CONCLUSION

DDSS currently still face a number of challenges, such as concerns about data protection and accuracy, and acceptance and awareness continue to be rather low. On the other hand, there is great potential for faster diagnosis, especially for rare diseases, which are easily overlooked due to their large number and the low awareness of them. The use of DDSS should therefore be carefully considered by doctors on a case-by-case basis.

The underuse of AI in the health sector: Opportunity costs, success stories, risks and recommendations

Purpose

This contribution explores the underuse of artificial intelligence (AI) in the health sector, what this means for practice, and how much the underuse can cost. Attention is drawn to the relevance of an issue that the European Parliament has outlined as a "major threat" in 2020. At its heart is the risk that research and development on trusted AI systems for medicine and digital health will pile up in lab centers without generating further practical relevance. Our analysis highlights why researchers, practitioners and especially policymakers, should pay attention to this phenomenon.

Methods

The paper examines the ways in which governments and public agencies are addressing the underuse of AI. As governments and international organizations often acknowledge the limitations of their own initiatives, the contribution explores the causes of the current issues and suggests ways to improve initiatives for digital health.

Results

Recommendations address the development of standards, models of regulatory governance, assessment of the opportunity costs of underuse of technology, and the urgency of the problem.

Conclusions

The exponential pace of AI advances and innovations makes the risks of underuse of AI increasingly threatening.

Graphical Abstract

Assessment of a Digital Symptom Checker Tool's Accuracy in Suggesting Reproductive Health Conditions: Clinical Vignettes Study

BACKGROUND

Reproductive health conditions such as endometriosis, uterine fibroids, and polycystic ovary syndrome (PCOS) affect a large proportion of women and people who menstruate worldwide. Prevalence estimates for these conditions range from 5% to 40% of women of reproductive age. Long diagnostic delays, up to 12 years, are common and contribute to health complications and increased health care costs. Symptom checker apps provide users with information and tools to better understand their symptoms and thus have the potential to reduce the time to diagnosis for reproductive health conditions.

OBJECTIVE

This study aimed to evaluate the agreement between clinicians and 3 symptom checkers (developed by Flo Health UK Limited) in assessing symptoms of endometriosis, uterine fibroids, and PCOS using vignettes. We also aimed to present a robust example of vignette case creation, review, and classification in the context of predeployment testing and validation of digital health symptom checker tools.

METHODS

Independent general practitioners were recruited to create clinical case vignettes of simulated users for the purpose of testing each condition symptom checker; vignettes created for each condition contained a mixture of condition-positive and condition-negative outcomes. A second panel of general practitioners then reviewed, approved, and modified (if necessary) each vignette. A third group of general practitioners reviewed each vignette case and designated a final classification. Vignettes were then entered into the symptom checkers by a fourth, different group of general practitioners. The outcomes of each symptom checker were then compared with the final classification of each vignette to produce accuracy metrics including percent agreement, sensitivity, specificity, positive predictive value, and negative predictive value.

RESULTS

A total of 24 cases were created per condition. Overall, exact matches between the vignette general practitioner classification and the symptom checker outcome were 83% (n=20) for endometriosis, 83% (n=20) for uterine fibroids, and 88% (n=21) for PCOS. For each symptom checker, sensitivity was reported as 81.8% for endometriosis, 84.6% for uterine fibroids, and 100% for PCOS; specificity was reported as 84.6% for endometriosis, 81.8% for uterine fibroids, and 75% for PCOS; positive predictive value was reported as 81.8% for endometriosis, 84.6% for uterine fibroids, 80% for PCOS; and negative predictive value was reported as 84.6% for endometriosis, 81.8% for uterine fibroids, and 100% for PCOS.

CONCLUSIONS

The single-condition symptom checkers have high levels of agreement with general practitioner classification for endometriosis, uterine fibroids, and PCOS. Given long delays in diagnosis for many reproductive health conditions, which lead to increased medical costs and potential health complications for individuals and health care providers, innovative health apps and symptom checkers hold the potential to improve care pathways.

Is language an issue? Accuracy of the German computerized diagnostic decision support system ISABEL and cross-validation with the English counterpart

OBJECTIVES

Existing computerized diagnostic decision support tools (CDDS) accurately return possible differential diagnoses (DDx) based on the clinical information provided. The German versions of the CDDS tools for clinicians (Isabel Pro) and patients (Isabel Symptom Checker) from ISABEL Healthcare have not been validated yet.

METHODS

We entered clinical features of 50 patient vignettes taken from an emergency medical text book and 50 real cases with a confirmed diagnosis derived from the electronic health record (EHR) of a large academic Swiss emergency room into the German versions of Isabel Pro and Isabel Symptom Checker. We analysed the proportion of DDx lists that included the correct diagnosis.

RESULTS

Isabel Pro and Symptom Checker provided the correct diagnosis in 82 and 71 % of the cases, respectively. Overall, the correct diagnosis was ranked in 71 , 61 and 37 % of the cases within the top 20, 10 and 3 of the provided DDx when using Isabel Pro. In general, accuracy was higher with vignettes than ED cases, i.e. listed the correct diagnosis more often (non-significant) and ranked the diagnosis significantly more often within the top 20, 10 and 3. On average, 38 ± 4.5 DDx were provided by Isabel Pro and Symptom Checker.

CONCLUSIONS

The German versions of Isabel achieved a somewhat lower accuracy compared to previous studies of the English version. The accuracy decreases substantially when the position in the suggested DDx list is taken into account. Whether Isabel Pro is accurate enough to improve diagnostic quality in clinical ED routine needs further investigation.

ChatGPT and most frequent urological diseases: analysing the quality of information and potential risks for patients

PURPOSE

Artificial intelligence (AI) is a set of systems or combinations of algorithms, which mimic human intelligence. ChatGPT is software with artificial intelligence which was recently developed by OpenAI. One of its potential uses could be to consult the information about pathologies and treatments. Our objective was to assess the quality of the information provided by AI like ChatGPT and establish if it is a secure source of information for patients.

METHODS

Questions about bladder cancer, prostate cancer, renal cancer, benign prostatic hypertrophy (BPH), and urinary stones were queried through ChatGPT 4.0. Two urologists analysed the responses provided by ChatGPT using DISCERN questionary and a brief instrument for evaluating the quality of informed consent documents.

RESULTS

The overall information provided in all pathologies was well-balanced. In each pathology was explained its anatomical location, affected population and a description of the symptoms. It concluded with the established risk factors and possible treatment. All treatment answers had a moderate quality score with DISCERN (3 of 5 points). The answers about surgical options contain the recovery time, type of anaesthesia, and potential complications. After analysing all the responses related to each disease, all pathologies except BPH achieved a DISCERN score of 4.

CONCLUSIONS

ChatGPT information should be used with caution since the chatbot does not disclose the sources of information and may contain bias even with simple questions related to the basics of urologic diseases.

Comparison of Diagnostic and Triage Accuracy of Ada Health and WebMD Symptom Checkers, ChatGPT, and Physicians for Patients in an Emergency Department: Clinical Data Analysis Study

BACKGROUND

Diagnosis is a core component of effective health care, but misdiagnosis is common and can put patients at risk. Diagnostic decision support systems can play a role in improving diagnosis by physicians and other health care workers. Symptom checkers (SCs) have been designed to improve diagnosis and triage (ie, which level of care to seek) by patients.

OBJECTIVE

The aim of this study was to evaluate the performance of the new large language model ChatGPT (versions 3.5 and 4.0), the widely used WebMD SC, and an SC developed by Ada Health in the diagnosis and triage of patients with urgent or emergent clinical problems compared with the final emergency department (ED) diagnoses and physician reviews.

METHODS

We used previously collected, deidentified, self-report data from 40 patients presenting to an ED for care who used the Ada SC to record their symptoms prior to seeing the ED physician. Deidentified data were entered into ChatGPT versions 3.5 and 4.0 and WebMD by a research assistant blinded to diagnoses and triage. Diagnoses from all 4 systems were compared with the previously abstracted final diagnoses in the ED as well as with diagnoses and triage recommendations from three independent board-certified ED physicians who had blindly reviewed the self-report clinical data from Ada. Diagnostic accuracy was calculated as the proportion of the diagnoses from ChatGPT, Ada SC, WebMD SC, and the independent physicians that matched at least one ED diagnosis (stratified as top 1 or top 3). Triage accuracy was calculated as the number of recommendations from ChatGPT, WebMD, or Ada that agreed with at least 2 of the independent physicians or were rated "unsafe" or "too cautious."

RESULTS

Overall, 30 and 37 cases had sufficient data for diagnostic and triage analysis, respectively. The rate of top-1 diagnosis matches for Ada, ChatGPT 3.5, ChatGPT 4.0, and WebMD was 9 (30%), 12 (40%), 10 (33%), and 12 (40%), respectively, with a mean rate of 47% for the physicians. The rate of top-3 diagnostic matches for Ada, ChatGPT 3.5, ChatGPT 4.0, and WebMD was 19 (63%), 19 (63%), 15 (50%), and 17 (57%), respectively, with a mean rate of 69% for physicians. The distribution of triage results for Ada was 62% (n=23) agree, 14% unsafe (n=5), and 24% (n=9) too cautious; that for ChatGPT 3.5 was 59% (n=22) agree, 41% (n=15) unsafe, and 0% (n=0) too cautious; that for ChatGPT 4.0 was 76% (n=28) agree, 22% (n=8) unsafe, and 3% (n=1) too cautious; and that for WebMD was 70% (n=26) agree, 19% (n=7) unsafe, and 11% (n=4) too cautious. The unsafe triage rate for ChatGPT 3.5 (41%) was significantly higher (P=.009) than that of Ada (14%).

CONCLUSIONS

ChatGPT 3.5 had high diagnostic accuracy but a high unsafe triage rate. ChatGPT 4.0 had the poorest diagnostic accuracy, but a lower unsafe triage rate and the highest triage agreement with the physicians. The Ada and WebMD SCs performed better overall than ChatGPT. Unsupervised patient use of ChatGPT for diagnosis and triage is not recommended without improvements to triage accuracy and extensive clinical evaluation.

Supporting primary care through symptom checking artificial intelligence: a study of patient and physician attitudes in Italian general practice

BACKGROUND

Rapid advancements in artificial intelligence (AI) have led to the adoption of AI-driven symptom checkers in primary care. This study aimed to evaluate both patients' and physicians' attitudes towards these tools in Italian general practice settings, focusing on their perceived utility, user satisfaction, and potential challenges.

METHODS

This feasibility study involved ten general practitioners (GPs) and patients visiting GP offices. The patients used a chatbot-based symptom checker before their medical visit and conducted anamnestic screening for COVID-19 and a medical history algorithm concerning the current medical problem. The entered data were forwarded to the GP as medical history aid. After the medical visit, both physicians and patients evaluated their respective symptoms. Additionally, physicians performed a final overall evaluation of the symptom checker after the conclusion of the practice phase.

RESULTS

Most patients did not use symptom checkers. Overall, 49% of patients and 27% of physicians reported being rather or very satisfied with the symptom checker. The most frequent patient-reported reasons for satisfaction were ease of use, precise and comprehensive questions, perceived time-saving potential, and encouragement of self-reflection. Every other patient would consider at-home use of the symptom checker for the first appraisal of health problems to save time, reduce unnecessary visits, and/or as an aid for the physician. Patients' attitudes towards the symptom checker were not significantly associated with age, sex, or level of education. Most patients (75%) and physicians (84%) indicated that the symptom checker had no effect on the duration of the medical visit. Only a few participants found the use of the symptom checker to be disruptive to the medical visit or its quality.

CONCLUSIONS

The findings suggest a positive reception of the symptom checker, albeit with differing focus between patients and physicians. With the potential to be integrated further into primary care, these tools require meticulous clinical guidance to maximize their benefits.

TRIAL REGISTRATION

The study was not registered, as it did not include direct medical intervention on human participants.

Diagnostic Performance, Triage Safety, and Usability of a Clinical Decision Support System Within a University Hospital Emergency Department: Algorithm Performance and Usability Study

Background

Computerized clinical decision support systems (CDSSs) are increasingly adopted in health care to optimize resources and streamline patient flow. However, they often lack scientific validation against standard medical care.

Objective

The purpose of this study was to assess the performance, safety, and usability of a CDSS in a university hospital emergency department setting in Kuopio, Finland.

Methods

Patients entering the emergency department were asked to voluntarily participate in this study. Patients aged 17 years or younger, patients with cognitive impairments, and patients who entered the unit in an ambulance or with the need for immediate care were excluded. Patients completed the CDSS web-based form and usability questionnaire when waiting for the triage nurse's evaluation. The CDSS data were anonymized and did not affect the patients' usual evaluation or treatment. Retrospectively, 2 medical doctors evaluated the urgency of each patient's condition by using the triage nurse's information, and urgent and nonurgent groups were created. The International Statistical Classification of Diseases, Tenth Revision diagnoses were collected from the electronic health records. Usability was assessed by using a positive version of the System Usability Scale questionnaire.

Results

In total, our analyses included 248 patients. Regarding urgency, the mean sensitivities were 85% and 19%, respectively, for urgent and nonurgent cases when assessing the performance of CDSS evaluations in comparison to that of physicians. The mean sensitivities were 85% and 35%, respectively, when comparing the evaluations between the two physicians. Our CDSS did not miss any cases that were evaluated to be emergencies by physicians; thus, all emergency cases evaluated by physicians were evaluated as either urgent cases or emergency cases by the CDSS. In differential diagnosis, the CDSS had an exact match accuracy of 45.5% (97/213). The usability was good, with a mean System Usability Scale score of 78.2 (SD 16.8).

Conclusions

In a university hospital emergency department setting with a large real-world population, our CDSS was found to be equally as sensitive in urgent patient cases as physicians and was found to have an acceptable differential diagnosis accuracy, with good usability. These results suggest that this CDSS can be safely assessed further in a real-world setting. A CDSS could accelerate triage by providing patient-provided data in advance of patients' initial consultations and categorize patient cases as urgent and nonurgent cases upon patients' arrival to the emergency department.

Triage and Diagnostic Accuracy of Online Symptom Checkers: Systematic Review

BACKGROUND

In the context of a deepening global shortage of health workers and, in particular, the COVID-19 pandemic, there is growing international interest in, and use of, online symptom checkers (OSCs). However, the evidence surrounding the triage and diagnostic accuracy of these tools remains inconclusive.

OBJECTIVE

This systematic review aimed to summarize the existing peer-reviewed literature evaluating the triage accuracy (directing users to appropriate services based on their presenting symptoms) and diagnostic accuracy of OSCs aimed at lay users for general health concerns.

METHODS

Searches were conducted in MEDLINE, Embase, CINAHL, Health Management Information Consortium (HMIC), and Web of Science, as well as the citations of the studies selected for full-text screening. We included peer-reviewed studies published in English between January 1, 2010, and February 16, 2022, with a controlled and quantitative assessment of either or both triage and diagnostic accuracy of OSCs directed at lay users. We excluded tools supporting health care professionals, as well as disease- or specialty-specific OSCs. Screening and data extraction were carried out independently by 2 reviewers for each study. We performed a descriptive narrative synthesis.

RESULTS

A total of 21,296 studies were identified, of which 14 (0.07%) were included. The included studies used clinical vignettes, medical records, or direct input by patients. Of the 14 studies, 6 (43%) reported on triage and diagnostic accuracy, 7 (50%) focused on triage accuracy, and 1 (7%) focused on diagnostic accuracy. These outcomes were assessed based on the diagnostic and triage recommendations attached to the vignette in the case of vignette studies or on those provided by nurses or general practitioners, including through face-to-face and telephone consultations. Both diagnostic accuracy and triage accuracy varied greatly among OSCs. Overall diagnostic accuracy was deemed to be low and was almost always lower than that of the comparator. Similarly, most of the studies (9/13, 69 %) showed suboptimal triage accuracy overall, with a few exceptions (4/13, 31%). The main variables affecting the levels of diagnostic and triage accuracy were the severity and urgency of the condition, the use of artificial intelligence algorithms, and demographic questions. However, the impact of each variable differed across tools and studies, making it difficult to draw any solid conclusions. All included studies had at least one area with unclear risk of bias according to the revised Quality Assessment of Diagnostic Accuracy Studies-2 tool.

CONCLUSIONS

Although OSCs have potential to provide accessible and accurate health advice and triage recommendations to users, more research is needed to validate their triage and diagnostic accuracy before widescale adoption in community and health care settings. Future studies should aim to use a common methodology and agreed standard for evaluation to facilitate objective benchmarking and validation.

TRIAL REGISTRATION

PROSPERO CRD42020215210; https://tinyurl.com/3949zw83.

Ethics and governance of trustworthy medical artificial intelligence

BACKGROUND

The growing application of artificial intelligence (AI) in healthcare has brought technological breakthroughs to traditional diagnosis and treatment, but it is accompanied by many risks and challenges. These adverse effects are also seen as ethical issues and affect trustworthiness in medical AI and need to be managed through identification, prognosis and monitoring.

METHODS

We adopted a multidisciplinary approach and summarized five subjects that influence the trustworthiness of medical AI: data quality, algorithmic bias, opacity, safety and security, and responsibility attribution, and discussed these factors from the perspectives of technology, law, and healthcare stakeholders and institutions. The ethical framework of ethical values-ethical principles-ethical norms is used to propose corresponding ethical governance countermeasures for trustworthy medical AI from the ethical, legal, and regulatory aspects.

RESULTS

Medical data are primarily unstructured, lacking uniform and standardized annotation, and data quality will directly affect the quality of medical AI algorithm models. Algorithmic bias can affect AI clinical predictions and exacerbate health disparities. The opacity of algorithms affects patients' and doctors' trust in medical AI, and algorithmic errors or security vulnerabilities can pose significant risks and harm to patients. The involvement of medical AI in clinical practices may threaten doctors 'and patients' autonomy and dignity. When accidents occur with medical AI, the responsibility attribution is not clear. All these factors affect people's trust in medical AI.

CONCLUSIONS

In order to make medical AI trustworthy, at the ethical level, the ethical value orientation of promoting human health should first and foremost be considered as the top-level design. At the legal level, current medical AI does not have moral status and humans remain the duty bearers. At the regulatory level, strengthening data quality management, improving algorithm transparency and traceability to reduce algorithm bias, and regulating and reviewing the whole process of the AI industry to control risks are proposed. It is also necessary to encourage multiple parties to discuss and assess AI risks and social impacts, and to strengthen international cooperation and communication.

How suitable are clinical vignettes for the evaluation of symptom checker apps? A test theoretical perspective

Objective

To evaluate the ability of case vignettes to assess the performance of symptom checker applications and to suggest refinements to the methodology used in case vignette-based audit studies.

Methods

We re-analyzed the publicly available data of two prominent case vignette-based symptom checker audit studies by calculating common metrics of test theory. Furthermore, we developed a new metric, the Capability Comparison Score (CCS), which compares symptom checker capability while controlling for the difficulty of the set of cases each symptom checker evaluated. We then scrutinized whether applying test theory and the CCS altered the performance ranking of the investigated symptom checkers.

Results

In both studies, most symptom checkers changed their rank order when adjusting the triage capability for item difficulty (ID) with the CCS. The previously reported triage accuracies commonly overestimated the capability of symptom checkers because they did not account for the fact that symptom checkers tend to selectively appraise easier cases (i.e., with high ID values). Also, many case vignettes in both studies showed insufficient (very low and even negative) values of item-total correlation (ITC), suggesting that individual items or the composition of item sets are of low quality.

Conclusions

A test-theoretic perspective helps identify previously undetected threats to the validity of case vignette-based symptom checker assessments and provides guidance and specific metrics to improve the quality of case vignettes, in particular by controlling for the difficulty of the vignettes an app was (not) able to evaluate correctly. Such measures might prove more meaningful than accuracy alone for the competitive assessment of symptom checkers. Our approach helps elaborate and standardize the methodology used for appraising symptom checker capability, which, ultimately, may yield more reliable results.

Challenges in evaluating the accuracy of AI-containing digital triage systems: A systematic review

INTRODUCTION

Patient-operated digital triage systems with AI components are becoming increasingly common. However, previous reviews have found a limited amount of research on such systems' accuracy. This systematic review of the literature aimed to identify the main challenges in determining the accuracy of patient-operated digital AI-based triage systems.

METHODS

A systematic review was designed and conducted in accordance with PRISMA guidelines in October 2021 using PubMed, Scopus and Web of Science. Articles were included if they assessed the accuracy of a patient-operated digital triage system that had an AI-component and could triage a general primary care population. Limitations and other pertinent data were extracted, synthesized and analysed. Risk of bias was not analysed as this review studied the included articles' limitations (rather than results). Results were synthesized qualitatively using a thematic analysis.

RESULTS

The search generated 76 articles and following exclusion 8 articles (6 primary articles and 2 reviews) were included in the analysis. Articles' limitations were synthesized into three groups: epistemological, ontological and methodological limitations. Limitations varied with regards to intractability and the level to which they can be addressed through methodological choices. Certain methodological limitations related to testing triage systems using vignettes can be addressed through methodological adjustments, whereas epistemological and ontological limitations require that readers of such studies appraise the studies with limitations in mind.

DISCUSSION

The reviewed literature highlights recurring limitations and challenges in studying the accuracy of patient-operated digital triage systems with AI components. Some of these challenges can be addressed through methodology whereas others are intrinsic to the area of inquiry and involve unavoidable trade-offs. Future studies should take these limitations in consideration in order to better address the current knowledge gaps in the literature.

Evidence synthesis, digital scribes, and translational challenges for artificial intelligence in healthcare

Healthcare has well-known challenges with safety, quality, and effectiveness, and many see artificial intelligence (AI) as essential to any solution. Emerging applications include the automated synthesis of best-practice research evidence including systematic reviews, which would ultimately see all clinical trial data published in a computational form for immediate synthesis. Digital scribes embed themselves in the process of care to detect, record, and summarize events and conversations for the electronic record. However, three persistent translational challenges must be addressed before AI is widely deployed. First, little effort is spent replicating AI trials, exposing patients to risks of methodological error and biases. Next, there is little reporting of patient harms from trials. Finally, AI built using machine learning may perform less effectively in different clinical settings.

Ethical, legal, and social aspects of symptom checker applications: a scoping review

Symptom Checker Applications (SCA) are mobile applications often designed for the end-user to assist with symptom assessment and self-triage. SCA are meant to provide the user with easily accessible information about their own health conditions. However, SCA raise questions regarding ethical, legal, and social aspects (ELSA), for example, regarding fair access to this new technology. The aim of this scoping review is to identify the ELSA of SCA in the scientific literature. A scoping review was conducted to identify the ELSA of SCA. Ten databases (e.g., Web of Science and PubMed) were used. Studies on SCA that address ELSA, written in English or German, were included in the review. The ELSA of SCA were extracted and synthesized using qualitative content analysis. A total of 25,061 references were identified, of which 39 were included in the analysis. The identified aspects were allotted to three main categories: (1) Technology; (2) Individual Level; and (3) Healthcare system. The results show that there are controversial debates in the literature on the ethical and social challenges of SCA usage. Furthermore, the debates are characterised by a lack of a specific legal perspective and empirical data. The review provides an overview on the spectrum of ELSA regarding SCA. It offers guidance to stakeholders in the healthcare system, for example, patients, healthcare professionals, and insurance providers and could be used in future empirical research to investigate the perspectives of those affected, such as users.

Assessing data gathering of chatbot based symptom checkers - a clinical vignettes study

BACKGROUND

The burden on healthcare systems is mounting continuously owing to population growth and aging, overuse of medical services, and the recent COVID-19 pandemic. This overload is also causing reduced healthcare quality and outcomes. One solution gaining momentum is the integration of intelligent self-assessment tools, known as symptom-checkers, into healthcare-providers' systems. To the best of our knowledge, no study so far has investigated the data-gathering capabilities of these tools, which represent a crucial resource for simulating doctors' skills in medical-interviews.

OBJECTIVES

The goal of this study was to evaluate the data-gathering function of currently available chatbot symptom-checkers.

METHODS

We evaluated 8 symptom-checkers using 28 clinical vignettes from the repository of MSD-Manual case studies. The mean number of predefined pertinent findings for each case was 31.8 ± 6.8. The vignettes were entered into the platforms by 3 medical students who simulated the role of the patient. For each conversation, we obtained the number of pertinent findings retrieved and the number of questions asked. We then calculated the recall-rates (pertinent-findings retrieved out of all predefined pertinent-findings), and efficiency-rates (pertinent-findings retrieved out of the number of questions asked) of data-gathering, and compared them between the platforms.

RESULTS

The overall recall rate for all symptom-checkers was 0.32(2,280/7,112;95 %CI 0.31-0.33) for all pertinent findings, 0.37(1,110/2,992;95 %CI 0.35-0.39) for present findings, and 0.28(1140/4120;95 %CI 0.26-0.29) for absent findings. Among the symptom-checkers, Kahun platform had the highest recall rate with 0.51(450/889;95 %CI 0.47-0.54). Out of 4,877 questions asked overall, 2,280 findings were gathered, yielding an efficiency rate of 0.46(95 %CI 0.45-0.48) across all platforms. Kahun was the most efficient tool 0.74 (95 %CI 0.70-0.77) without a statistically significant difference from Your.MD 0.69(95 %CI 0.65-0.73).

CONCLUSION

The data-gathering performance of currently available symptom checkers is questionable. From among the tools available, Kahun demonstrated the best overall performance.

Online Symptom Checkers: Recommendations for a Vignette-Based Clinical Evaluation Standard

The use of patient-facing online symptom checkers (OSCs) has expanded in recent years, but their accuracy, safety, and impact on patient behaviors and health care systems remain unclear. The lack of a standardized process of clinical evaluation has resulted in significant variation in approaches to OSC validation and evaluation. The aim of this paper is to characterize a set of congruent requirements for a standardized vignette-based clinical evaluation process of OSCs. Discrepancies in the findings of comparative studies to date suggest that different steps in OSC evaluation methodology can significantly influence outcomes. A standardized process with a clear specification for vignette-based clinical evaluation is urgently needed to guide developers and facilitate the objective comparison of OSCs. We propose 15 recommendation requirements for an OSC evaluation standard. A third-party evaluation process and protocols for prospective real-world evidence studies should also be prioritized to quality assure OSC assessment.

MediCoSpace : Visual Decision-Support for Doctor-Patient Consultations using Medical Concept Spaces from EHRs

Healthcare systems are under pressure from an aging population, rising costs, and increasingly complex conditions and treatments. Although data are determined to play a bigger role in how doctors diagnose and prescribe treatments, they struggle due to a lack of time and an abundance of structured and unstructured information. To address this challenge, we introduce MediCoSpace , a visual decision-support tool for more efficient doctor-patient consultations. The tool links patient reports to past and present diagnoses, diseases, drugs, and treatments, both for the current patient and other patients in comparable situations. MediCoSpace uses textual medical data, deep-learning supported text analysis and concept spaces to facilitate a visual discovery process. The tool is evaluated with five medical doctors. The results show that MediCoSpace facilitates a promising, yet complex way to discover unlikely relations and thus suggests a path toward the development of interactive visual tools to provide physicians with more holistic diagnoses and personalized, dynamic treatments for patients.

Patients' self-triage for unscheduled urgent care: a preliminary study on the accuracy and factors affecting the performance of a Belgian self-triage platform

Background

Management of unscheduled urgent care is a complex concern for many healthcare providers. Facing the challenge of appropriately dispatching unscheduled care, primary and emergency physicians have collaboratively implemented innovative strategies such as telephone triage. Currently, new original solutions tend to emerge with the development of new technologies. We created an interactive patient self-triage platform, ODISSEE, and aimed to explore its accuracy and potential factors affecting its performance using clinical case scenarios.

Methods

The ODISSEE platform was developed based on previously validated triage protocols for out-of-hours primary care. ODISSEE is composed of 18 icons leading to algorithmic questions that finally provide an advised orientation (emergency or primary care services). To investigate ODISSEE performance, we used 100 clinical case scenarios, each associated with a preestablished orientation determined by a group of experts. Fifteen volunteers were asked to self-triage with 50 randomly selected scenarios using ODISSEE on a digital tablet. Their triage results were compared with the experts’ references.

Results

The 15 participants performed a total of 750 self-triages, which matched the experts references regarding the level of care in 85.6% of the cases. The orientation was incorrect in 14.4%, with an undertriage rate of 1.9% and an overtriage rate of 12.5%. The tool’s specificity and sensitivity to advise participants on the appropriate level of care were 69% (95% CI: 64—74) and 97% (95% CI: 95—98) respectively. When combined with advice on the level of urgency, the tool only found the correct orientation in 68.4% with 9.2% of undertriages and 22.4% of overtriages. Some participant characteristics and the types of medical conditions demonstrated a significant association with the tool performance.

Conclusion

Self-triage apps, such as the ODISSEE platform, could represent an innovative method to allow patients to self-triage to the most appropriate level of care. This study based on clinical vignettes highlights some positive arguments regarding ODISSEE safety, but further research is needed to assess the generalizability of such tools to the population without equity issues.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-022-08571-5.

Learning from the Failure of Autonomous and Intelligent Systems: Accidents, Safety, and Sociotechnical Sources of Risk

Efforts to develop autonomous and intelligent systems (AIS) have exploded across a range of settings in recent years, from self-driving cars to medical diagnostic chatbots. These have the potential to bring enormous benefits to society but also have the potential to introduce new-or amplify existing-risks. As these emerging technologies become more widespread, one of the most critical risk management challenges is to ensure that failures of AIS can be rigorously analyzed and understood so that the safety of these systems can be effectively governed and improved. AIS are necessarily developed and deployed within complex human, social, and organizational systems, but to date there has been little systematic examination of the sociotechnical sources of risk and failure in AIS. Accordingly, this article develops a conceptual framework that characterizes key sociotechnical sources of risk in AIS by reanalyzing one of the most publicly reported failures to date: the 2018 fatal crash of Uber's self-driving car. Publicly available investigative reports were systematically analyzed using constant comparative analysis to identify key sources and patterns of sociotechnical risk. Five fundamental domains of sociotechnical risk were conceptualized-structural, organizational, technological, epistemic, and cultural-each indicated by particular patterns of sociotechnical failure. The resulting SOTEC framework of sociotechnical risk in AIS extends existing theories of risk in complex systems and highlights important practical and theoretical implications for managing risk and developing infrastructures of learning in AIS.

The diagnostic and triage accuracy of digital and online symptom checker tools: a systematic review

Digital and online symptom checkers are an increasingly adopted class of health technologies that enable patients to input their symptoms and biodata to produce a set of likely diagnoses and associated triage advice. However, concerns regarding the accuracy and safety of these symptom checkers have been raised. This systematic review evaluates the accuracy of symptom checkers in providing diagnoses and appropriate triage advice. MEDLINE and Web of Science were searched for studies that used either real or simulated patients to evaluate online or digital symptom checkers. The primary outcomes were the diagnostic and triage accuracy of the symptom checkers. The QUADAS-2 tool was used to assess study quality. Of the 177 studies retrieved, 10 studies met the inclusion criteria. Researchers evaluated the accuracy of symptom checkers using a variety of medical conditions, including ophthalmological conditions, inflammatory arthritides and HIV. A total of 50% of the studies recruited real patients, while the remainder used simulated cases. The diagnostic accuracy of the primary diagnosis was low across included studies (range: 19–37.9%) and varied between individual symptom checkers, despite consistent symptom data input. Triage accuracy (range: 48.8–90.1%) was typically higher than diagnostic accuracy. Overall, the diagnostic and triage accuracy of symptom checkers are variable and of low accuracy. Given the increasing push towards adopting this class of technologies across numerous health systems, this study demonstrates that reliance upon symptom checkers could pose significant patient safety hazards. Large-scale primary studies, based upon real-world data, are warranted to demonstrate the adequate performance of these technologies in a manner that is non-inferior to current best practices. Moreover, an urgent assessment of how these systems are regulated and implemented is required.

Impact of COVID-19 in mental health trusts

Objective

To explore how mental health trusts in England adapted and responded to the challenges posed by the COVID-19 pandemic, with the aim of identifying lessons that can be learned during and beyond the pandemic.

Methods

Following a scoping study, we undertook 52 semi-structured interviews with senior managers, clinicians, patient representatives and commissioning staff across four case study sites. These sites varied in size, location and grading awarded by a national regulatory body. We explored how services have been repurposed and reorganized in response to the pandemic and the participants’ perceptions of the impact of these changes on quality of care and the wellbeing of staff.

Results

Mental health trusts have shown great flexibility and resilience in rapidly implementing new models of care and developing creative digital solutions at speed. New collaborative arrangements have been stimulated by a shared sense of urgency and enabled by additional funding and a more permissive policy environment. But there has also been a significant negative impact on the wellbeing of staff, particularly those staff from a minority ethnic background. Also, there were concerns that digital technology could effectively disenfranchise some vulnerable groups and exacerbate existing health inequalities.

Conclusions

Many of the service changes and digital innovations undertaken during the pandemic appear promising. Nevertheless, those changes need to be urgently and rigorously appraised to assure their effectiveness and to assess their impact on social exclusion and health inequalities.

User initiated symptom assessment with an electronic symptom checker. Study protocol for mixed-methods validation. (Preprint)

BACKGROUND

The national Omaolo digital social welfare and health care service of Finland provides a symptom checker, Omaolo, which is a medical device (based on Duodecim Clinical Decision Support EBMEDS software) with a CE marking (risk class IIa), manufactured by the government-owned DigiFinland Oy. Users of this service can perform their triage by using the questions in the symptom checker. By completing the symptom checker, the user receives a recommendation for action and a service assessment with appropriate guidance regarding their health problems on the basis of a selected specific symptom in the symptom checker. This allows users to be provided with appropriate health care services, regardless of time and place.

OBJECTIVE

This study describes the protocol for the mixed methods validation process of the symptom checker available in Omaolo digital services.

METHODS

This is a mixed methods study using quantitative and qualitative methods, which will be part of the clinical validation process that takes place in primary health care centers in Finland. Each organization provides a space where the study and the nurse triage can be done in order to include an unscreened target population of users. The primary health care units provide walk-in model services, where no prior phone call or contact is required. For the validation of the Omaolo symptom checker, case vignettes will be incorporated to supplement the triage accuracy of rare and acute cases that cannot be tested extensively in real-life settings. Vignettes are produced from a variety of clinical sources, and they test the symptom checker in different triage levels by using 1 standardized patient case example.

RESULTS

This study plan underwent an ethics review by the regional permission, which was requested from each organization participating in the research, and an ethics committee statement was requested and granted from Pirkanmaa hospital district's ethics committee, which is in accordance with the University of Tampere's regulations. Of 964 clinical user-filled symptom checker assessments, 877 cases were fully completed with a triage result, and therefore, they met the requirements for clinical validation studies. The goal for sufficient data has been reached for most of the chief symptoms. Data collection was completed in September 2019, and the first feasibility and patient experience results were published by the end of 2020. Case vignettes have been identified and are to be completed before further testing the symptom checker. The analysis and reporting are estimated to be finalized in 2024.

CONCLUSIONS

The primary goals of this multimethod electronic symptom checker study are to assess safety and to provide crucial information regarding the accuracy and usability of the Omaolo electronic symptom checker. To our knowledge, this will be the first study to include real-life clinical cases along with case vignettes.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

DERR1-10.2196/41423.

Assessing the Clinical Robustness of Digital Health Startups: Cross-sectional Observational Analysis

BACKGROUND

The digital health sector has experienced rapid growth over the past decade. However, health care technology stakeholders lack a comprehensive understanding of clinical robustness and claims across the industry.

OBJECTIVE

This analysis aimed to examine the clinical robustness and public claims made by digital health companies.

METHODS

A cross-sectional observational analysis was conducted using company data from the Rock Health Digital Health Venture Funding Database, the US Food and Drug Administration, and the US National Library of Medicine. Companies were included if they sell products targeting the prevention, diagnosis, or treatment phases of the care continuum. Clinical robustness was defined using regulatory filings and clinical trials completed by each company. Public claims data included clinical, economic, and engagement claims regarding product outcomes made by each company on its website.

RESULTS

A total of 224 digital health companies with an average age of 7.7 years were included in our cohort. Average clinical robustness was 2.5 (1.8 clinical trials and 0.8 regulatory filings) with a median score of 1. Ninety-eight (44%) companies had a clinical robustness score of 0, while 45 (20%) companies had a clinical robustness score of 5 or more. The average number of public claims was 1.3 (0.5 clinical, 0.4 economic, and 0.4 engagement); the median number of claims was 1. No correlation was observed between clinical robustness and number of clinical claims (r2=0.02), clinical robustness and total funding (r2=0.08), or clinical robustness and company age (r2=0.18).

CONCLUSIONS

Many digital health companies have a low level of clinical robustness and do not make many claims as measured by regulatory filings, clinical trials, and public data shared online. Companies and customers may benefit from investing in greater clinical validation efforts.

Triage Accuracy of Symptom Checker Apps: 5-Year Follow-up Evaluation

Background

Symptom checkers are digital tools assisting laypersons in self-assessing the urgency and potential causes of their medical complaints. They are widely used but face concerns from both patients and health care professionals, especially regarding their accuracy. A 2015 landmark study substantiated these concerns using case vignettes to demonstrate that symptom checkers commonly err in their triage assessment.

Objective

This study aims to revisit the landmark index study to investigate whether and how symptom checkers’ capabilities have evolved since 2015 and how they currently compare with laypersons’ stand-alone triage appraisal.

Methods

In early 2020, we searched for smartphone and web-based applications providing triage advice. We evaluated these apps on the same 45 case vignettes as the index study. Using descriptive statistics, we compared our findings with those of the index study and with publicly available data on laypersons’ triage capability.

Results

We retrieved 22 symptom checkers providing triage advice. The median triage accuracy in 2020 (55.8%, IQR 15.1%) was close to that in 2015 (59.1%, IQR 15.5%). The apps in 2020 were less risk averse (odds 1.11:1, the ratio of overtriage errors to undertriage errors) than those in 2015 (odds 2.82:1), missing >40% of emergencies. Few apps outperformed laypersons in either deciding whether emergency care was required or whether self-care was sufficient. No apps outperformed the laypersons on both decisions.

Conclusions

Triage performance of symptom checkers has, on average, not improved over the course of 5 years. It decreased in 2 use cases (advice on when emergency care is required and when no health care is needed for the moment). However, triage capability varies widely within the sample of symptom checkers. Whether it is beneficial to seek advice from symptom checkers depends on the app chosen and on the specific question to be answered. Future research should develop resources (eg, case vignette repositories) to audit the capabilities of symptom checkers continuously and independently and provide guidance on when and to whom they should be recommended.

Patient's Perception of Digital Symptom Assessment Technologies in Rheumatology: Results From a Multicentre Study

Introduction

An increasing number of digital tools, including dedicated diagnostic decision support systems (DDSS) exist to better assess new symptoms and understand when and where to seek medical care. The aim of this study was to evaluate patient's previous online assessment experiences and to compare the acceptability, usability, usefulness and potential impact of artificial intelligence (AI)-based symptom checker (Ada) and an online questionnaire-based self-referral tool (Rheport).

Materials and Methods

Patients newly presenting to three German secondary rheumatology outpatient clinics were randomly assigned in a 1:1 ratio to complete consecutively Ada or Rheport in a prospective non-blinded multicentre controlled crossover randomized trial. DDSS completion time was recorded by local study personnel and perceptions on DDSS and previous online assessment were collected through a self-completed study questionnaire, including usability measured with the validated System Usability Scale (SUS).

Results

600 patients (median age 52 years, 418 women) were included. 277/600 (46.2%) of patients used an online search engine prior to the appointment. The median time patients spent assessing symptoms was 180, 7, and 8 min, respectively using online using search engines, Ada and Rheport. 111/275 (40.4%), 266/600 (44.3%) and 395/600 (65.8%) of patients rated the respective symptom assessment as very helpful or helpful, using online search engines, Ada and Rheport, respectively. Usability of both diagnostic decision support systems (DDSS) was “good” with a significantly higher mean SUS score (SD) of Rheport 77.1/100 (16.0) compared to Ada 74.4/100 (16.8), (p < 0.0001). In male patients, usability of Rheport was rated higher than Ada (p = 0.02) and the usability rating of older (52 years ≥) patients of both DDSS was lower than in younger participants (p = 0.005). Both effects were independent of each other. 440/600 (73.3%) and 475/600 (79.2%) of the patients would recommend Ada and Rheport to friends and other patients, respectively.

Conclusion

In summary, patients increasingly assess their symptoms independently online, however only a minority used dedicated symptom assessment websites or DDSS. DDSS, such as Ada an Rheport are easy to use, well accepted among patients with musculoskeletal complaints and could replace online search engines for patient symptom assessment, potentially saving time and increasing helpfulness.

What is the suitability of clinical vignettes in benchmarking the performance of online symptom checkers? An audit study

OBJECTIVE

Assess the suitability of clinical vignettes in benchmarking the performance of online symptom checkers (OSCs).

DESIGN

Observational study using a publicly available free OSC.

PARTICIPANTS

Healthily OSC, which provided consultations in English, was used to record consultation outcomes from two lay and four expert inputters using 139 standardised patient vignettes. Each vignette included three diagnostic solutions and a triage recommendation in one of three categories of triage urgency. A panel of three independent general practitioners interpreted the vignettes to arrive at an alternative set of diagnostic and triage solutions. Both sets of diagnostic and triage solutions were consolidated to arrive at a final consolidated version for benchmarking.

MAIN OUTCOME MEASURES

Six inputters simulated 834 standardised patient evaluations using Healthily OSC and recorded outputs (triage solution, signposting, and whether the correct diagnostic solution appeared first or within the first three differentials). We estimated Cohen's kappa to assess how interpretations by different inputters could lead to divergent OSC output even when using the same vignette or when compared with a separate panel of physicians.

RESULTS

There was moderate agreement on triage recommendation (kappa=0.48), and substantial agreement on consultation outcomes between all inputters (kappa=0.73). OSC performance improved significantly from baseline when compared against the final consolidated diagnostic and triage solution (p<0.001).

CONCLUSIONS

Clinical vignettes are inherently limited in their utility to benchmark the diagnostic accuracy or triage safety of OSC. Real-world evidence studies involving real patients are recommended to benchmark the performance of OSC against a panel of physicians.

Study protocol for a pilot prospective, observational study investigating the condition suggestion and urgency advice accuracy of a symptom assessment app in sub-Saharan Africa: the AFYA-'Health' Study

INTRODUCTION

Due to a global shortage of healthcare workers, there is a lack of basic healthcare for 4 billion people worldwide, particularly affecting low-income and middle-income countries. The utilisation of AI-based healthcare tools such as symptom assessment applications (SAAs) has the potential to reduce the burden on healthcare systems. The purpose of the AFYA Study (AI-based Assessment oF health sYmptoms in TAnzania) is to evaluate the accuracy of the condition suggestions and urgency advice provided by a user on a Swahili language Ada SAA.

METHODS AND ANALYSIS

This study is designed as an observational prospective clinical study. The setting is a waiting room of a Tanzanian district hospital. It will include patients entering the outpatient clinic with various conditions and age groups, including children and adolescents. Patients will be asked to use the SAA before proceeding to usual care. After usual care, they will have a consultation with a study-provided physician. Patients and healthcare practitioners will be blinded to the SAA's results. An expert panel will compare the Ada SAA's condition suggestions and urgency advice to usual care and study provided differential diagnoses and triage. The primary outcome measures are the accuracy and comprehensiveness of the Ada SAA evaluated against the gold standard differential diagnoses.

ETHICS AND DISSEMINATION

Ethical approval was received by the ethics committee (EC) of Muhimbili University of Health and Allied Sciences with an approval number MUHAS-REC-09-2019-044 and the National Institute for Medical Research, NIMR/HQ/R.8c/Vol. I/922. All amendments to the protocol are reported and adapted on the basis of the requirements of the EC. The results from this study will be submitted to peer-reviewed journals, local and international stakeholders, and will be communicated in editorials/articles by Ada Health.

TRIAL REGISTRATION NUMBER

NCT04958577.

The Good, the Bad, and the Invisible with Its Opportunity Costs: Introduction to the ‘J’ Special Issue on “the Impact of Artificial Intelligence on Law”

Scholars and institutions have been increasingly debating the moral and legal challenges of AI, together with the models of governance that should strike the balance between the opportunities and threats brought forth by AI, its ‘good’ and ‘bad’ facets. There are more than a hundred declarations on the ethics of AI and recent proposals for AI regulation, such as the European Commission’s AI Act, have further multiplied the debate. Still, a normative challenge of AI is mostly overlooked, and regards the underuse, rather than the misuse or overuse, of AI from a legal viewpoint. From health care to environmental protection, from agriculture to transportation, there are many instances of how the whole set of benefits and promises of AI can be missed or exploited far below its full potential, and for the wrong reasons: business disincentives and greed among data keepers, bureaucracy and professional reluctance, or public distrust in the era of no-vax conspiracies theories. The opportunity costs that follow this technological underuse is almost terra incognita due to the ‘invisibility’ of the phenomenon, which includes the ‘shadow prices’ of economy. This introduction provides metrics for such assessment and relates this work to the development of new standards for the field. We must quantify how much it costs not to use AI systems for the wrong reasons.

Diagnostic Performance of an App-Based Symptom Checker in Mental Disorders: Comparative Study in Psychotherapy Outpatients

BACKGROUND

Digital technologies have become a common starting point for health-related information-seeking. Web- or app-based symptom checkers aim to provide rapid and accurate condition suggestions and triage advice but have not yet been investigated for mental disorders in routine health care settings.

OBJECTIVE

This study aims to test the diagnostic performance of a widely available symptom checker in the context of formal diagnosis of mental disorders when compared with therapists' diagnoses based on structured clinical interviews.

METHODS

Adult patients from an outpatient psychotherapy clinic used the app-based symptom checker Ada-check your health (ADA; Ada Health GmbH) at intake. Accuracy was assessed as the agreement of the first and 1 of the first 5 condition suggestions of ADA with at least one of the interview-based therapist diagnoses. In addition, sensitivity, specificity, and interrater reliabilities (Gwet first-order agreement coefficient [AC1]) were calculated for the 3 most prevalent disorder categories. Self-reported usability (assessed using the System Usability Scale) and acceptance of ADA (assessed using an adapted feedback questionnaire) were evaluated.

RESULTS

A total of 49 patients (30/49, 61% women; mean age 33.41, SD 12.79 years) were included in this study. Across all patients, the interview-based diagnoses matched ADA's first condition suggestion in 51% (25/49; 95% CI 37.5-64.4) of cases and 1 of the first 5 condition suggestions in 69% (34/49; 95% CI 55.4-80.6) of cases. Within the main disorder categories, the accuracy of ADA's first condition suggestion was 0.82 for somatoform and associated disorders, 0.65 for affective disorders, and 0.53 for anxiety disorders. Interrater reliabilities ranged from low (AC1=0.15 for anxiety disorders) to good (AC1=0.76 for somatoform and associated disorders). The usability of ADA was rated as high in the System Usability Scale (mean 81.51, SD 11.82, score range 0-100). Approximately 71% (35/49) of participants would have preferred a face-to-face over an app-based diagnostic.

CONCLUSIONS

Overall, our findings suggest that a widely available symptom checker used in the formal diagnosis of mental disorders could provide clinicians with a list of condition suggestions with moderate-to-good accuracy. However, diagnostic performance was heterogeneous between disorder categories and included low interrater reliability. Although symptom checkers have some potential to complement the diagnostic process as a screening tool, the diagnostic performance should be tested in larger samples and in comparison with further diagnostic instruments.

Agreement and Reliability Analysis of Machine Learning Scaling and Wireless Monitoring in the Assessment of Acute Proximal Weakness by Experts and Non-Experts: A Feasibility Study

Assessing the symptoms of proximal weakness caused by neurological deficits requires the knowledge and experience of neurologists. Recent advances in machine learning and the Internet of Things have resulted in the development of automated systems that emulate physicians’ assessments. The application of those systems requires not only accuracy in the classification but also reliability regardless of users’ proficiency in the real environment for the clinical point-of-care and the personalized health management. This study provides an agreement and reliability analysis of using a machine learning-based scaling of Medical Research Council (MRC) proximal scores to evaluate proximal weakness by experts and non-experts. The system trains an ensemble learning model using the signals from sensors attached to the limbs of patients in a neurological intensive care unit. For the agreement analysis, we investigated the percent agreement of MRC proximal scores and Bland-Altman plots of kinematic features between the expert- and non-expert scaling. We also analyzed the intra-class correlation coefficients (ICCs) of kinematic features and Krippendorff’s alpha of the observers’ scaling for the reliability analysis. The mean percent agreement between the expert- and the non-expert scaling was 0.542 for manual scaling and 0.708 for autonomous scaling. The ICCs of kinematic features measured using sensors ranged from 0.742 to 0.850, whereas the Krippendorff’s alpha of manual scaling for the three observers was 0.275. The autonomous assessment system can be utilized by the caregivers, paramedics, or other observers during an emergency to evaluate acute stroke patients.

Performance of a new symptom checker in patient triage: Canadian cohort study

BACKGROUND

Computerized algorithms known as symptom checkers aim to help patients decide what to do should they have a new medical concern. However, despite widespread implementation, most studies on symptom checkers have involved simulated patients. Only limited evidence currently exists about symptom checker safety or accuracy when used by real patients. We developed a new prototype symptom checker and assessed its safety and accuracy in a prospective cohort of patients presenting to primary care and emergency departments with new medical concerns.

METHOD

A prospective cohort study was done to assess the prototype's performance. The cohort consisted of adult patients (≥16 years old) who presented to hospital emergency departments and family physician clinics. Primary outcomes were safety and accuracy of triage recommendations to seek hospital care, seek primary care, or manage symptoms at home.

RESULTS

Data from 281 hospital patients and 300 clinic patients were collected and analyzed. Sensitivity to emergencies was 100% (10/10 encounters). Sensitivity to urgencies was 90% (73/81) and 97% (34/35) for hospital and primary care patients, respectively. The prototype was significantly more accurate than patients at triage (73% versus 58%, p<0.01). Compliance with triage recommendations in this cohort using this iteration of the symptom checker would have reduced hospital visits by 55% but cause potential harm in 2-3% from delay in care.

INTERPRETATION

The prototype symptom checker was superior to patients in deciding the most appropriate treatment setting for medical issues. This symptom checker could reduce a significant number of unnecessary hospital visits, with accuracy and safety outcomes comparable to existing data on telephone triage.

Quality of condition suggestions and urgency advice provided by the Ada symptom assessment app evaluated with vignettes optimised for Australia

When people face a health problem, they often first ask, 'Is there an app for that?'. We investigated the quality of advice provided by the Ada symptom assessment application to address the question, 'How do I know the app on my phone is safe and provides good advice?'. The app was tested with 48 independently created vignettes developed for a previous study, including 18 specifically developed for the Australian setting, using an independently developed methodology to evaluate the accuracy of condition suggestions and urgency advice. The correct condition was listed first in 65% of vignettes, and in the Top 3 results in 83% of vignettes. The urgency advice in the app exactly matched the gold standard 63% of vignettes. The app's accuracy of condition suggestion and urgency advice is higher than that of the best-performing symptom assessment app reported in a previous study (61%, 77% and 52% for conditions suggested in the Top 1, Top 3 and exactly matching urgency advice respectively). These results are relevant to the application of symptom assessment in primary and community health, where medical quality and safety should determine app choice.

Automated identification of diagnostic labelling errors in medicine

Objectives

Identification of diagnostic error is complex and mostly relies on expert ratings, a severely limited procedure. We developed a system that allows to automatically identify diagnostic labelling error from diagnoses coded according to the international classification of diseases (ICD), often available as routine health care data.

Methods

The system developed (index test) was validated against rater based classifications taken from three previous studies of diagnostic labeling error (reference standard). The system compares pairs of diagnoses through calculation of their distance within the ICD taxonomy. Calculation is based on four different algorithms. To assess the concordance between index test and reference standard, we calculated the area under the receiver operating characteristics curve (AUROC) and corresponding confidence intervals. Analysis were conducted overall and separately per algorithm and type of available dataset.

Results

Diagnoses of 1,127 cases were analyzed. Raters previously classified 24.58% of cases as diagnostic labelling errors (ranging from 12.3 to 87.2% in the three datasets). AUROC ranged between 0.821 and 0.837 overall, depending on the algorithm used to calculate the index test (95% CIs ranging from 0.8 to 0.86). Analyzed per type of dataset separately, the highest AUROC was 0.924 (95% CI 0.887–0.962).

Conclusions

The trigger system to automatically identify diagnostic labeling error from routine health care data performs excellent, and is unaffected by the reference standards’ limitations. It is however only applicable to cases with pairs of diagnoses, of which one must be more accurate or otherwise superior than the other, reflecting a prevalent definition of a diagnostic labeling error.

SERIES: eHealth in primary care. Part 5: A critical appraisal of five widely used eHealth applications for primary care - opportunities and challenges

Background

Given the pressure on modern healthcare systems, eHealth can offer valuable opportunities. However, understanding the potential and challenges of eHealth in daily practice can be challenging for many general practitioners (GPs) and their staff.

Objectives

To critically appraise five widely used eHealth applications, in relation to safe, evidence-based and high-quality eHealth. Using these applications as examples, we aim to increase understanding of eHealth among GPs and highlight the opportunities and challenges presented by eHealth.

Discussion

eHealth applications can support patients while increasing efficiency for GPs. A three-way division (inform, monitor, track; interaction; data utilisation) characterises many eHealth applications, with an increasing degree of complexity depending on the domain. All applications provide information and some have extra functionalities that promote interaction, while data analysis and artificial intelligence may be applied to support or (fully) automate care processes. Applications in the inform domain are relatively easy to use and implement but their impact on clinical outcomes may be limited. More demanding applications, in terms of privacy and ethical aspects, are found in the data utilisation domain and may potentially have a more significant impact on care processes and patient outcomes. When selecting and implementing eHealth applications, we recommend that GPs remain critical regarding preconditions on safe, evidence-based and high-quality eHealth, particularly in the case of more complex applications in the data utilisation domain.

Clinical Decision Support Systems for Diagnosis in Primary Care: A Scoping Review

Diagnosis is one of the crucial tasks performed by primary care physicians; however, primary care is at high risk of diagnostic errors due to the characteristics and uncertainties associated with the field. Prevention of diagnostic errors in primary care requires urgent action, and one of the possible methods is the use of health information technology. Its modes such as clinical decision support systems (CDSS) have been demonstrated to improve the quality of care in a variety of medical settings, including hospitals and primary care centers, though its usefulness in the diagnostic domain is still unknown. We conducted a scoping review to confirm the usefulness of the CDSS in the diagnostic domain in primary care and to identify areas that need to be explored. Search terms were chosen to cover the three dimensions of interest: decision support systems, diagnosis, and primary care. A total of 26 studies were included in the review. As a result, we found that the CDSS and reminder tools have significant effects on screening for common chronic diseases; however, the CDSS has not yet been fully validated for the diagnosis of acute and uncommon chronic diseases. Moreover, there were few studies involving non-physicians.

Accuracy of online symptom checkers and the potential impact on service utilisation

OBJECTIVES

The aims of our study are firstly to investigate the diagnostic and triage performance of symptom checkers, secondly to assess their potential impact on healthcare utilisation and thirdly to investigate for variation in performance between systems.

SETTING

Publicly available symptom checkers for patient use.

PARTICIPANTS

Publicly available symptom-checkers were identified. A standardised set of 50 clinical vignettes were developed and systematically run through each system by a non-clinical researcher.

PRIMARY AND SECONDARY OUTCOME MEASURES

System accuracy was assessed by measuring the percentage of times the correct diagnosis was a) listed first, b) within the top five diagnoses listed and c) listed at all. The safety of the disposition advice was assessed by comparing it with national guidelines for each vignette.

RESULTS

Twelve tools were identified and included. Mean diagnostic accuracy of the systems was poor, with the correct diagnosis being present in the top five diagnoses on 51.0% (Range 22.2 to 84.0%). Safety of disposition advice decreased with condition urgency (being 71.8% for emergency cases vs 87.3% for non-urgent cases). 51.0% of systems suggested additional resource utilisation above that recommended by national guidelines (range 18.0% to 61.2%). Both diagnostic accuracy and appropriate resource recommendation varied substantially between systems.

CONCLUSIONS

There is wide variation in performance between available symptom checkers and overall performance is significantly below what would be accepted in any other medical field, though some do achieve a good level of accuracy and safety of disposition. External validation and regulation are urgently required to ensure these public facing tools are safe.

Use of symptom checkers for COVID-19-related symptoms among university students: a qualitative study

OBJECTIVE

Symptom checkers are potentially beneficial tools during pandemics. To increase the use of the platform, perspectives of end users must be gathered. Our objectives were to understand the perspectives and experiences of young adults related to the use of symptom checkers for assessing COVID-19-related symptoms and to identify areas for improvement.

METHODS

We conducted semistructured qualitative interviews with 22 young adults (18-34 years of age) at a university in Ontario, Canada. Interviews were audio-recorded, transcribed, and analysed using inductive thematic analysis.

RESULTS

We identified six main themes related to the decision of using a symptom checker for COVID-19 symptoms: (1) presence of symptoms or a combination of symptoms, (2) knowledge about COVID-19 symptoms, (3) fear of seeking in-person healthcare services, (4) awareness about symptom checkers, (5) paranoia and (6) curiosity. Participants who used symptom checkers shared by governmental entities reported an overall positive experience. Individuals who used non-credible sources reported suboptimal experiences due to lack of perceived credibility. Five main areas for improvement were identified: (1) information about the creators of the platform, (2) explanation of symptoms, (3) personalised experience, (4) language options, and (5) option to get tested.

CONCLUSIONS

This study suggests an increased acceptance of symptom checkers due to the perceived risks of infection associated with seeking in-person healthcare services. Symptom checkers have the potential to reduce the burden on healthcare systems and health professionals, especially during pandemics; however, these platforms could be improved to increase use.

Benchmarking Triage Capability of Symptom Checkers Against That of Medical Laypersons: Survey Study

BACKGROUND

Symptom checkers (SCs) are tools developed to provide clinical decision support to laypersons. Apart from suggesting probable diagnoses, they commonly advise when users should seek care (triage advice). SCs have become increasingly popular despite prior studies rating their performance as mediocre. To date, it is unclear whether SCs can triage better than those who might choose to use them.

OBJECTIVE

This study aims to compare triage accuracy between SCs and their potential users (ie, laypersons).

METHODS

On Amazon Mechanical Turk, we recruited 91 adults from the United States who had no professional medical background. In a web-based survey, the participants evaluated 45 fictitious clinical case vignettes. Data for 15 SCs that had processed the same vignettes were obtained from a previous study. As main outcome measures, we assessed the accuracy of the triage assessments made by participants and SCs for each of the three triage levels (ie, emergency care, nonemergency care, self-care) and overall, the proportion of participants outperforming each SC in terms of accuracy, and the risk aversion of participants and SCs by comparing the proportion of cases that were overtriaged.

RESULTS

The mean overall triage accuracy was similar for participants (60.9%, SD 6.8%; 95% CI 59.5%-62.3%) and SCs (58%, SD 12.8%). Most participants outperformed all but 5 SCs. On average, SCs more reliably detected emergencies (80.6%, SD 17.9%) than laypersons did (67.5%, SD 16.4%; 95% CI 64.1%-70.8%). Although both SCs and participants struggled with cases requiring self-care (the least urgent triage category), SCs more often wrongly classified these cases as emergencies (43/174, 24.7%) compared with laypersons (56/1365, 4.10%).

CONCLUSIONS

Most SCs had no greater triage capability than an average layperson, although the triage accuracy of the five best SCs was superior to the accuracy of most participants. SCs might improve early detection of emergencies but might also needlessly increase resource utilization in health care. Laypersons sometimes require support in deciding when to rely on self-care but it is in that very situation where SCs perform the worst. Further research is needed to determine how to best combine the strengths of humans and SCs.

Safety concerns with consumer-facing mobile health applications and their consequences: a scoping review

Objective

To summarize the research literature about safety concerns with consumer-facing health apps and their consequences.

Materials and Methods

We searched bibliographic databases including PubMed, Web of Science, Scopus, and Cochrane libraries from January 2013 to May 2019 for articles about health apps. Descriptive information about safety concerns and consequences were extracted and classified into natural categories. The review was conducted in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) statement.

Results

Of the 74 studies identified, the majority were reviews of a single or a group of similar apps (n = 66, 89%), nearly half related to disease management (n = 34, 46%). A total of 80 safety concerns were identified, 67 related to the quality of information presented including incorrect or incomplete information, variation in content, and incorrect or inappropriate response to consumer needs. The remaining 13 related to app functionality including gaps in features, lack of validation for user input, delayed processing, failure to respond to health dangers, and faulty alarms. Of the 52 reports of actual or potential consequences, 5 had potential for patient harm. We also identified 66 reports about gaps in app development, including the lack of expert involvement, poor evidence base, and poor validation.

Conclusions

Safety of apps is an emerging public health issue. The available evidence shows that apps pose clinical risks to consumers. Involvement of consumers, regulators, and healthcare professionals in development and testing can improve quality. Additionally, mandatory reporting of safety concerns is needed to improve outcomes.

[Brave New Psychiatry? Or: What Astray is Artificial Intelligence Leading Psychiatry?]

The article summarizes various publications on the application of "learning algorithms" and "artificial neural networks" in psychiatry to describe a dystopian future scenario. The draft of a nosology based on molecular biology is opposed to the ecological disturbance concept of psychiatry developed from the critical examination of history and in dialogue with the stakeholders.

Assessment of Diagnosis and Triage in Validated Case Vignettes Among Nonphysicians Before and After Internet Search

IMPORTANCE

When confronted with new medical symptoms, many people turn to the internet to understand why they are ill as well as whether and where they should get care. Such searches may be harmful because they may facilitate misdiagnosis and inappropriate triage.

OBJECTIVE

To empirically measure the association of an internet search for health information with diagnosis, triage, and anxiety by laypeople.

DESIGN, SETTING, AND PARTICIPANTS

This survey study used a nationally representative sample of US adults who were recruited through an online platform between April 1, 2019, and April 15, 2019. A total of 48 validated case vignettes of both common (eg, viral illness) and severe (eg, heart attack) conditions were used. Participants were asked to relay their diagnosis, triage, and anxiety regarding 1 of these cases before and after searching the internet for health information.

EXPOSURES

Short, validated case vignettes written at or below the sixth-grade reading level randomly assigned to participants.

MAIN OUTCOMES AND MEASURES

Correct diagnosis, correct triage, and flipping (changing) or anchoring (not changing) diagnosis and triage decisions were the main outcomes. Multivariable modeling was performed to identify patient factors associated with correct triage and diagnosis.

RESULTS

Of the 5000 participants, 2549 were female (51.0%), 3819 were White (76.4%), and the mean (SD) age was 45.0 (16.9) years. Mean internet search time was 12.1 (95% CI, 10.7-13.5) minutes per case. No difference in triage accuracy was found before and after search (74.5% vs 74.1%; difference, -0.4 [95% CI, -1.4 to 0.6]; P = .06), but improved diagnostic accuracy was found (49.8% vs 54.0%; difference, 4.2% [95% CI, 3.1%-5.3%]; P < .001). Most participants (4254 [85.1%]) were anchored on their diagnosis. Of the 14.9% of participants (n = 746) who flipped their diagnosis, 9.6% (n = 478) flipped from incorrect to correct and 5.4% (n = 268) flipped from correct to incorrect. The following groups had an increased rate of correct diagnosis: adults 40 years or older (eg, 40-49 years: 5.1 [95% CI, 0.8-9.4] percentage points better than those aged <30 years; P = .02), women (9.4 [95% CI, 6.8-12.0] percentage points better than men; P < .001), and those with perceived poor health status (16.3 [95% CI, 6.9-25.6] percentage points better than those with excellent status; P = .001) and with more than 2 chronic diseases (6.8 [95% CI, 1.5-12.1] percentage points better than those with 0 conditions; P = .01).

CONCLUSIONS AND RELEVANCE

This study found that an internet search for health information was associated with small increases in diagnostic accuracy but not with triage accuracy.

Teasing out Artificial Intelligence in Medicine: An Ethical Critique of Artificial Intelligence and Machine Learning in Medicine

The rapid adoption and implementation of artificial intelligence in medicine creates an ontologically distinct situation from prior care models. There are both potential advantages and disadvantages with such technology in advancing the interests of patients, with resultant ontological and epistemic concerns for physicians and patients relating to the instatiation of AI as a dependent, semi- or fully-autonomous agent in the encounter. The concept of libertarian paternalism potentially exercised by AI (and those who control it) has created challenges to conventional assessments of patient and physician autonomy. The unclear legal relationship between AI and its users cannot be settled presently, an progress in AI and its implementation in patient care will necessitate an iterative discourse to preserve humanitarian concerns in future models of care. This paper proposes that physicians should neither uncritically accept nor unreasonably resist developments in AI but must actively engage and contribute to the discourse, since AI will affect their roles and the nature of their work. One's moral imaginative capacity must be engaged in the questions of beneficence, autonomy, and justice of AI and whether its integration in healthcare has the potential to augment or interfere with the ends of medical practice.