Eating disorders early app use mediates treatment effect on clinical improvement.

OBJECTIVE: Eating disorders (EDs) contribute considerably to the global burden of disease. However, most affected individuals do not receive treatment. Mobile apps present an enormous opportunity to increase access to mental healthcare services. This study examined whether the degree of usage of a self-help app for EDs mediated the app's effects on the clinical response by individuals with EDs. METHOD: App usage measures included the total number of cognitive-behavioral meal logs, total number of days spent using the app, and the last day the app was used during the study period. Mediation analysis was performed using the MacArthur framework. RESULTS: All usage variables met the analytic requirements for testing mediation (group means (sd) for app and standard app, respectively: logs = 74 (108) vs. 51.4 (88.1), days spent = 14.3 (17.5) vs. 10.6 (15.0), p-values from Wilcox rank sum tests p < .01). Regression coefficients indicated mediation effects. The mediation effects demonstrated support that increased engagement (as measured by logs and time spent on the app) was related to an increased likelihood of achieving a significant clinical change by the end of the trial. DISCUSSION: Greater and longer engagement in an ED app mediates its efficacy in terms of ED remission.

Subgrouping the users of a specialized app for eating disorders.

Although mobile technologies for eating disorders (EDs) are burgeoning, there is limited data about the clinical characteristics of individuals using specialized smartphone applications (apps) without accompanying traditional forms of treatment. This study evaluated whether the users of an ED app cluster in clinically meaningful groups. Participants were 1,280 app users (91.3% female; mean age 27) who reported not being in a weekly treatment for their ED. A hierarchical cluster analysis distinguished five groups of participants, all approximating DSM-5 ED categories. One cluster comprised of non-female, ethnically diverse users with Bulimia Nervosa features. Findings suggest that app users resemble known patient classifications.

Accelerated failure time model under general biased sampling scheme.

Right-censored time-to-event data are sometimes observed from a (sub)cohort of patients whose survival times can be subject to outcome-dependent sampling schemes. In this paper, we propose a unified estimation method for semiparametric accelerated failure time models under general biased estimating schemes. The proposed estimator of the regression covariates is developed upon a bias-offsetting weighting scheme and is proved to be consistent and asymptotically normally distributed. Large sample properties for the estimator are also derived. Using rank-based monotone estimating functions for the regression parameters, we find that the estimating equations can be easily solved via convex optimization. The methods are confirmed through simulations and illustrated by application to real datasets on various sampling schemes including length-bias sampling, the case-cohort design and its variants.

Perceptions of the Professional Development Value of Honorary Fellowship Award Experiences.

OBJECTIVE: Professional societies engage in activities with the aim of nurturing highly talented early career members of their field. Little is known about the value of honorary fellowship awards given annually by professional societies. Following up on the only known prior study of this topic, authors queried fellowship awardees in one psychiatric society to better understand the perceived value of honorary fellowships and other outcomes, such as subsequent involvement in professional societies. METHODS: The authors queried former participants in the Laughlin and Psychiatry Resident-In-Training Examination® (PRITE®) Programs regarding their fellowship experiences and their subsequent involvement in The American College of Psychiatrists and other psychiatry membership organizations. The authors obtained frequency data and analyzed responses using t-tests and chi-squared tests. Associations between the outcomes and demographic characteristics such as age, gender, and fellowship type was tested. RESULTS: Responses were gathered from 143 individuals who had participated in the Laughlin Fellowship and 22 in the PRITE Fellowship. Respondents felt that that the fellowship experience had been helpful professionally. Laughlin fellows were older and more likely to have assumed a leadership role in professional organizations (60 % vs 36 %, p = 0.04). Laughlin fellows also more strongly endorsed professional recognition as a benefit at the time of receiving their award. Survey respondents reported increased participation in professional organizations and assumed leadership roles in The College and other professional organizations subsequent to the fellowship experience. CONCLUSIONS: On the whole, fellows were generally positive about their experiences. Many respondents became involved with The College subsequent to their fellowship, but a larger proportion became involved with other organizations, including in leadership roles. Professional societies with early career programs such as the Laughlin Fellowship and the PRITE Fellowship appear to identify and support future leaders as intended, but these leaders may engage more with other professional societies.

Informal health care practices of residents: "curbside" consultation and self-diagnosis and treatment.

OBJECTIVE: The authors explore the experiences of residents with respect to informal care related to personal health, including "curbside consultation," self-diagnosis, and self-prescription-self-care practices that run counter to ethical guidelines in medicine. METHODS: Residents at one medical school completed a written survey regarding their personal health care practices, including their experiences in seeking or providing informal consultation, self-diagnosis, and self-prescribing. The authors obtained frequency data and analyzed responses via cross-tabulation. They used logistic regression models to assess the association of reported informal care practices and potential confounders, such as age, gender, and training program. RESULTS: One hundred and fifty-five residents volunteered (71 % response rate). Most respondents had sought health care formally (70 %), and more had sought informal care in the previous 12 months (80 %). Of those who had pursued informal care, 90 % endorsed having requested a physical exam, a laboratory test, or a medication prescription from an attending, resident, or medical student. Respondents (28 %) commonly endorsed prescribing medication for themselves. Most respondents (90 %) reported being approached for informal care at least once in the previous year, including 84 % who were approached for prescriptions and 22 % who were approached by attending physicians. Main reasons endorsed for informal care seeking related to busy schedules and to cost and confidentiality advantages. Psychiatry residents reported using both formal and informal channels for personal health care, and 31 % acknowledged prescribing medications for themselves. CONCLUSIONS: Informal care was a common practice among the residents in this study. Residents reported time constraints as the greatest influence on informal care seeking, rather than reasons found in previous studies related to cost and quality of care, protection of confidentiality, or prevention of embarrassment. The effects of informal care practices involving resident physicians warrant additional study.

Understanding Physician's Perspectives on AI in Health Care: Protocol for a Sequential Multiple Assignment Randomized Vignette Study.

BACKGROUND: As the availability and performance of artificial intelligence (AI)-based clinical decision support (CDS) systems improve, physicians and other care providers poised to be on the front lines will be increasingly tasked with using these tools in patient care and incorporating their outputs into clinical decision-making processes. Vignette studies provide a means to explore emerging hypotheses regarding how context-specific factors, such as clinical risk, the amount of information provided about the AI, and the AI result, may impact physician acceptance and use of AI-based CDS tools. To best anticipate how such factors influence the decision-making of frontline physicians in clinical scenarios involving AI decision-support tools, hypothesis-driven research is needed that enables scenario testing before the implementation and deployment of these tools. OBJECTIVE: This study's objectives are to (1) design an original, web-based vignette-based survey that features hypothetical scenarios based on emerging or real-world applications of AI-based CDS systems that will vary systematically by features related to clinical risk, the amount of information provided about the AI, and the AI result; and (2) test and determine causal effects of specific factors on the judgments and perceptions salient to physicians' clinical decision-making. METHODS: US-based physicians with specialties in family or internal medicine will be recruited through email and mail (target n=420). Through a web-based survey, participants will be randomized to a 3-part "sequential multiple assignment randomization trial (SMART) vignette" detailing a hypothetical clinical scenario involving an AI decision support tool. The SMART vignette design is similar to the SMART design but adapted to a survey design. Each respondent will be randomly assigned to 1 of the possible vignette variations of the factors we are testing at each stage, which include the level of clinical risk, the amount of information provided about the AI, and the certainty of the AI output. Respondents will be given questions regarding their hypothetical decision-making in response to the hypothetical scenarios. RESULTS: The study is currently in progress and data collection is anticipated to be completed in 2024. CONCLUSIONS: The web-based vignette study will provide information on how contextual factors such as clinical risk, the amount of information provided about an AI tool, and the AI result influence physicians' reactions to hypothetical scenarios that are based on emerging applications of AI in frontline health care settings. Our newly proposed "SMART vignette" design offers several benefits not afforded by the extensively used traditional vignette design, due to the 2 aforementioned features. These advantages are (1) increased validity of analyses targeted at understanding the impact of a factor on the decision outcome, given previous outcomes and other contextual factors; and (2) balanced sample sizes across groups. This study will generate a better understanding of physician decision-making within this context. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/54787.

Academic machine learning researchers' ethical perspectives on algorithm development for health care: a qualitative study.

OBJECTIVES: We set out to describe academic machine learning (ML) researchers' ethical considerations regarding the development of ML tools intended for use in clinical care. MATERIALS AND METHODS: We conducted in-depth, semistructured interviews with a sample of ML researchers in medicine (N = 10) as part of a larger study investigating stakeholders' ethical considerations in the translation of ML tools in medicine. We used a qualitative descriptive design, applying conventional qualitative content analysis in order to allow participant perspectives to emerge directly from the data. RESULTS: Every participant viewed their algorithm development work as holding ethical significance. While participants shared positive attitudes toward continued ML innovation, they described concerns related to data sampling and labeling (eg, limitations to mitigating bias; ensuring the validity and integrity of data), and algorithm training and testing (eg, selecting quantitative targets; assessing reproducibility). Participants perceived a need to increase interdisciplinary training across stakeholders and to envision more coordinated and embedded approaches to addressing ethics issues. DISCUSSION AND CONCLUSION: Participants described key areas where increased support for ethics may be needed; technical challenges affecting clinical acceptability; and standards related to scientific integrity, beneficence, and justice that may be higher in medicine compared to other industries engaged in ML innovation. Our results help shed light on the perspectives of ML researchers in medicine regarding the range of ethical issues they encounter or anticipate in their work, including areas where more attention may be needed to support the successful development and integration of medical ML tools.

A note on Using regression models to analyze randomized trials: asymptotically valid hypothesis tests despite incorrectly specified models.

In the context of randomized trials, Rosenblum and van der Laan (2009, Biometrics 63, 937-945) considered the null hypothesis of no treatment effect on the mean outcome within strata of baseline variables. They showed that hypothesis tests based on linear regression models and generalized linear regression models are guaranteed to have asymptotically correct Type I error regardless of the actual data generating distribution, assuming the treatment assignment is independent of covariates. We consider another important outcome in randomized trials, the time from randomization until failure, and the null hypothesis of no treatment effect on the survivor function conditional on a set of baseline variables. By a direct application of arguments in Rosenblum and van der Laan (2009), we show that hypothesis tests based on multiplicative hazards models with an exponential link, i.e., proportional hazards models, and multiplicative hazards models with linear link functions where the baseline hazard is parameterized, are asymptotically valid under model misspecification provided that the censoring distribution is independent of the treatment assignment given the covariates. In the case of the Cox model and linear link model with unspecified baseline hazard function, the arguments in Rosenblum and van der Laan (2009) cannot be applied to show the robustness of a misspecified model. Instead, we adopt an approach used in previous literature (Struthers and Kalbfleisch, 1986, Biometrika 73, 363-369) to show that hypothesis tests based on these models, including models with interaction terms, have correct type I error.

Validation of reported predialysis nephrology care of older patients initiating dialysis.

The Centers for Medicare and Medicaid Services (CMS) Medical Evidence Report (form CMS-2728) queries providers about the timing of the patient's first nephrologist consultation before initiation of dialysis. The monitoring of disease-specific goals in the Healthy People 2020 initiative will use information from this question, but the accuracy of the reported information is unknown. We defined a cohort of 80,509 patients aged ≥67 years who initiated dialysis between July 2005 and December 2008 with ≥2 years of uninterrupted Medicare coverage as their primary payer. The primary referent, determined from claims data, was the first observed outpatient nephrologist consultation; secondary analyses used the earliest nephrology consultation, whether inpatient or outpatient. We used linear regression models to assess the associations among the magnitude of discrepant reporting and patient characteristics and we tested for any temporal trends. When using the earliest recorded outpatient nephrology encounter, agreement between the two sources of ascertainment was 48.2%, and the κ statistic was 0.29 when we categorized the timing of the visit into four periods (never, <6, 6-12, and >12 months). When we dichotomized the timing of first predialysis nephrology care at >12 or ≤12 months, accuracy was 70% (κ=0.36), but it differed by patient characteristics and declined over time. In conclusion, we found substantial disagreement between information from the CMS Medical Evidence Report and Medicare physician claims on the timing of first predialysis nephrologist care. More-specific instructions may improve reporting and increase the utility of form CMS-2728 for research and public health surveillance.

Physicians' and Machine Learning Researchers' Perspectives on Ethical Issues in the Early Development of Clinical Machine Learning Tools: Qualitative Interview Study.

BACKGROUND: Innovative tools leveraging artificial intelligence (AI) and machine learning (ML) are rapidly being developed for medicine, with new applications emerging in prediction, diagnosis, and treatment across a range of illnesses, patient populations, and clinical procedures. One barrier for successful innovation is the scarcity of research in the current literature seeking and analyzing the views of AI or ML researchers and physicians to support ethical guidance. OBJECTIVE: This study aims to describe, using a qualitative approach, the landscape of ethical issues that AI or ML researchers and physicians with professional exposure to AI or ML tools observe or anticipate in the development and use of AI and ML in medicine. METHODS: Semistructured interviews were used to facilitate in-depth, open-ended discussion, and a purposeful sampling technique was used to identify and recruit participants. We conducted 21 semistructured interviews with a purposeful sample of AI and ML researchers (n=10) and physicians (n=11). We asked interviewees about their views regarding ethical considerations related to the adoption of AI and ML in medicine. Interviews were transcribed and deidentified by members of our research team. Data analysis was guided by the principles of qualitative content analysis. This approach, in which transcribed data is broken down into descriptive units that are named and sorted based on their content, allows for the inductive emergence of codes directly from the data set. RESULTS: Notably, both researchers and physicians articulated concerns regarding how AI and ML innovations are shaped in their early development (ie, the problem formulation stage). Considerations encompassed the assessment of research priorities and motivations, clarity and centeredness of clinical needs, professional and demographic diversity of research teams, and interdisciplinary knowledge generation and collaboration. Phase-1 ethical issues identified by interviewees were notably interdisciplinary in nature and invited questions regarding how to align priorities and values across disciplines and ensure clinical value throughout the development and implementation of medical AI and ML. Relatedly, interviewees suggested interdisciplinary solutions to these issues, for example, more resources to support knowledge generation and collaboration between developers and physicians, engagement with a broader range of stakeholders, and efforts to increase diversity in research broadly and within individual teams. CONCLUSIONS: These qualitative findings help elucidate several ethical challenges anticipated or encountered in AI and ML for health care. Our study is unique in that its use of open-ended questions allowed interviewees to explore their sentiments and perspectives without overreliance on implicit assumptions about what AI and ML currently are or are not. This analysis, however, does not include the perspectives of other relevant stakeholder groups, such as patients, ethicists, industry researchers or representatives, or other health care professionals beyond physicians. Additional qualitative and quantitative research is needed to reproduce and build on these findings.

Influence of Dispositional Optimism on Ethically Salient Research Perspectives: A Pilot Study.

Research participants should be drawn as fairly as possible from the potential volunteer population. Underlying personality traits are underexplored as factors influencing research decision-making. Dispositional optimism, known to affect coping, physical health, and psychological well-being, has been minimally studied with respect to research-related attitudes. We conducted an exploratory, online survey with 151 individuals (with self-reported mental illness [n = 50], physical illness [n = 51], or neither [n = 50]) recruited via MTurk. We evaluated associations between dispositional optimism (assessed with the Life Orientation Test-Revised) and general research attitudes, perceived protectiveness of five research safeguards, and willingness to participate in research using safeguards. Strongly optimistic respondents expressed more positive research attitudes and perceived four safeguards as more positively influencing willingness to participate. Optimism was positively associated with expressed willingness to participate in clinical research. Our findings add to a limited literature on the influence of individual traits on ethically salient research perspectives.

Factors Influencing Perceived Helpfulness and Participation in Innovative Research: A Pilot Study of Individuals with and without Mood Symptoms.

Little is known about how individuals with and without mood disorders perceive the inherent risks and helpfulness of participating in innovative psychiatric research, or about the factors that influence their willingness to participate. We conducted an online survey with 80 individuals (self-reported mood disorder [n = 25], self-reported good health [n = 55]) recruited via MTurk. We assessed respondents' perceptions of risk and helpfulness in study vignettes associated with two innovative research projects (intravenous ketamine therapy and wearable devices), as well as their willingness to participate in these projects. Respondents with and without mood disorders perceived risk similarly across projects. Respondents with no mood disorders viewed both projects as more helpful to society than to research volunteers, while respondents with mood disorders viewed the projects as equally helpful to volunteers and society. Individuals with mood disorders perceived ketamine research, and the two projects on average, as more helpful to research volunteers than did individuals without mood disorders. Our findings add to a limited empirical literature on the perspectives of volunteers in innovative psychiatric research.

Self-reported influences on willingness to receive COVID-19 vaccines among physically ill, mentally ill, and healthy individuals.

OBJECTIVE: Individuals with mental and physical disorders have been disproportionately affected by adverse health outcomes due to the COVID-19 pandemic, and yet vaccine hesitancy persists despite clear evidence of health benefits. Therefore, our study explored factors influencing willingness to receive a COVID-19 vaccine. METHODS: Individuals with mental illness (n = 332), physical illness (n = 331), and no health issues (n = 328) were recruited via Amazon Mechanical Turk. Participants rated willingness to obtain a fully approved COVID-19 vaccine or a vaccine approved only for experimental/emergency use and influences in six domains upon their views. We examined differences by health status. RESULTS: Participants across groups were moderately willing to receive a COVID-19 vaccine. Perceived risk was negatively associated with willingness. Participants differentiated between vaccine risk by approval stage and were less willing to receive an experimental vaccine. Individuals with mental illness rated risk of both vaccines similarly to healthy individuals. Individuals with physical illness expressed less willingness to receive an experimental vaccine. Domain influences differently affected willingness by health status as well as by vaccine approval status. CONCLUSIONS: Our findings are reassuring regarding the ability of people with mental disorders to appreciate risk in medical decision-making and the ability of people of varied health backgrounds to distinguish between the benefits and risks of clinical care and research, refuting the prevailing notions of psychiatric exceptionalism and therapeutic misconception. Our findings shine a light on potential paths forward to support vaccine acceptance.

Perceived protectiveness of research safeguards and influences on willingness to participate in research: A novel MTurk pilot study.

Little is known about how individuals with mood disorders view the protectiveness of research safeguards, and whether their views affect their willingness to participate in psychiatric research. We conducted an online survey with 80 individuals (self-reported mood disorder [n = 25], self-reported good health [n = 55]) recruited via MTurk. We assessed respondents' perceptions of the protectiveness of five common research safeguards, as well as their willingness to participate in research that incorporates each safeguard. Perceived protectiveness was strongly related to willingness to participate in research for four of the safeguards. Our findings add to a limited literature on the motivations and perspectives of key stakeholders in psychiatric research.

Attitudes of Mothers Regarding Willingness to Enroll Their Children in Research.

This study assessed mothers' perspectives regarding research involvement by their children, factors that might affect perceptions of research risks, and attitudes regarding willingness to enroll children in research. Participants completed a survey on Amazon Mechanical Turk. Mothers were less inclined to enroll children in research involving procedures posing higher risk (regression coefficient = -0.51). Mothers without mental health issues with children without health issues were more sensitive to risk than mothers without mental health issues with children with health issues (estimated difference = 0.49). Mothers with mental health issues were more willing than mothers without mental health issues to enroll children in research (regression coefficient = -0.90). Among mothers with mental health issues, having a child with a health issue was associated with increased willingness to enroll in research, compared with having children without health issues (estimated difference = 0.65).

Negative association of perceived risk and willingness to participate in innovative psychiatric research protocols.

Psychiatric researchers grapple with concerns that individuals with mental illness may be less likely to appreciate risks of research participation, particularly compared to people not suffering from mental illness. Therefore, empirical studies that directly compare the perspectives of such individuals are needed. In addition, it is important to evaluate perspectives regarding varied types of research protocols, particularly as innovative psychiatric research protocols emerge. In this pilot study, respondents with a mood disorder (n = 25) as well as respondents without a mood disorder (n = 55) were recruited using Amazon's Mechanical Turk (MTurk) platform. These respondents were surveyed regarding four psychiatric research projects (i.e., experimental medication [pill form]; non-invasive magnetic brain stimulation; experimental medication [intravenous infusion]; and implantation of a device in the brain). Regardless of health status, respondents rated the four research protocols as somewhat to highly risky. The brain-device implant protocol was seen as the most risky, while the magnetic brain stimulation project was viewed as "somewhat risky". Respondents, on average and regardless of health status, rated their willingness at or below "somewhat willing." Respondents were least willing to participate in the brain-device implant protocol, whereas they were "somewhat willing" to participate in the magnetic brain stimulation protocol. Trust in medical research was negatively associated with perceived risk of research protocols. Perceived risk was negatively associated with willingness to participate, even when adjusting for potential confounders, suggesting that attunement to risk crosses diagnostic, gender, and ethnic categories, and is more salient to research decision-making than trust in medical research and dispositional optimism. The findings of this study may offer reassurance about the underlying decision-making processes of individuals considering participation in innovative neuroscience studies.

Artificial Intelligence and Suicide Prevention: A Systematic Review of Machine Learning Investigations.

Suicide is a leading cause of death that defies prediction and challenges prevention efforts worldwide. Artificial intelligence (AI) and machine learning (ML) have emerged as a means of investigating large datasets to enhance risk detection. A systematic review of ML investigations evaluating suicidal behaviors was conducted using PubMed/MEDLINE, PsychInfo, Web-of-Science, and EMBASE, employing search strings and MeSH terms relevant to suicide and AI. Databases were supplemented by hand-search techniques and Google Scholar. Inclusion criteria: (1) journal article, available in English, (2) original investigation, (3) employment of AI/ML, (4) evaluation of a suicide risk outcome. N = 594 records were identified based on abstract search, and 25 hand-searched reports. N = 461 reports remained after duplicates were removed, n = 316 were excluded after abstract screening. Of n = 149 full-text articles assessed for eligibility, n = 87 were included for quantitative synthesis, grouped according to suicide behavior outcome. Reports varied widely in methodology and outcomes. Results suggest high levels of risk classification accuracy (>90%) and Area Under the Curve (AUC) in the prediction of suicidal behaviors. We report key findings and central limitations in the use of AI/ML frameworks to guide additional research, which hold the potential to impact suicide on broad scale.