25. Multiple Imputation for Nonresponse in Surveys. By D. B. Rubin. ISBN 0 471 08705 X. Wiley, Chichester, 1987. 258 pp. £30.25.

Multiple Imputation for Nonresponse in Surveys

The use of artificial intelligence, and the deep-learning subtype in particular, has been enabled by the use of labeled big data, along with markedly enhanced computing power and cloud storage, across all sectors. In medicine, this is beginning to have an impact at three levels: for clinicians, predominantly via rapid, accurate image interpretation; for health systems, by improving workflow and the potential for reducing medical errors; and for patients, by enabling them to process their own data to promote health. The current limitations, including bias, privacy and security, and lack of transparency, along with the future directions of these applications will be discussed in this article. Over time, marked improvements in accuracy, productivity, and workflow will likely be actualized, but whether that will be used to improve the patient-doctor relationship or facilitate its erosion remains to be seen.

High-performance medicine: the convergence of human and artificial intelligence

Suicidal thoughts and behaviors (STBs) are major public health problems that have not declined appreciably in several decades. One of the first steps to improving the prevention and treatment of STBs is to establish risk factors (i.e., longitudinal predictors). To provide a summary of current knowledge about risk factors, we conducted a meta-analysis of studies that have attempted to longitudinally predict a specific STB-related outcome. This included 365 studies (3,428 total risk factor effect sizes) from the past 50 years. The present random-effects meta-analysis produced several unexpected findings: across odds ratio, hazard ratio, and diagnostic accuracy analyses, prediction was only slightly better than chance for all outcomes; no broad category or subcategory accurately predicted far above chance levels; predictive ability has not improved across 50 years of research; studies rarely examined the combined effect of multiple risk factors; risk factors have been homogenous over time, with 5 broad categories accounting for nearly 80% of all risk factor tests; and the average study was nearly 10 years long, but longer studies did not produce better prediction. The homogeneity of existing research means that the present meta-analysis could only speak to STB risk factor associations within very narrow methodological limits-limits that have not allowed for tests that approximate most STB theories. The present meta-analysis accordingly highlights several fundamental changes needed in future studies. In particular, these findings suggest the need for a shift in focus from risk factors to machine learning-based risk algorithms. (PsycINFO Database Record

Risk factors for suicidal thoughts and behaviors: A meta-analysis of 50 years of research.

There is increasing concern that most current published research findings are false. The probability that a research claim is true may depend on study power and bias, the number of other studies on the same question, and, importantly, the ratio of true to no relationships among the relationships probed in each scientific field. In this framework, a research finding is less likely to be true when the studies conducted in a field are smaller; when effect sizes are smaller; when there is a greater number and lesser preselection of tested relationships; where there is greater flexibility in designs, definitions, outcomes, and analytical modes; when there is greater financial and other interest and prejudice; and when more teams are involved in a scientific field in chase of statistical significance. Simulations show that for most study designs and settings, it is more likely for a research claim to be false than true. Moreover, for many current scientific fields, claimed research findings may often be simply accurate measures of the prevailing bias. In this essay, I discuss the implications of these problems for the conduct and interpretation of research.

Why most published research findings are false

Why People Die by Suicide

OBJECTIVE: Self-injurious thoughts and behaviors (SITBs) are a major public health problem that traditional interventions have been unable to address on a large scale. The goal of this series of studies was to take initial steps toward developing an effective SITB treatment that can be easily delivered on a very large scale. METHOD: We created a brief (1-2 min), game-like app called Therapeutic Evaluative Conditioning (TEC), designed to increase aversion to SITBs and decrease aversion to the self. In 3 separate studies, we recruited participants with recent and severe histories of SITBs from web forums focused on self-injury and psychopathology (Ns = 114, 131, and 163) and randomly assigned them to receive access to the mobile treatment TEC app or a control app for 1 month. We tested the effect of TEC on the frequency of self-cutting, nonsuicidal self-injury more generally, suicide ideation, suicide plans, and suicidal behaviors. RESULTS: Analyses showed that, compared with the control app, TEC produced moderate reductions for all SITBs except suicide ideation. Across studies, the largest and most consistent reductions were for self-cutting episodes (32%-40%), suicide plans (21%-59%), and suicidal behaviors (33%-77%). Two of the 3 studies showed that TEC impacted its intended treatment targets and that greater change in these targets was associated with greater SITB reductions. TEC effects were not maintained at the 1-month posttreatment follow-up. CONCLUSIONS: Future versions of brief, mobile interventions like that tested here may have the potential to reduce SITBs and related behaviors on a large scale. (PsycINFO Database Record(c) 2016 APA, all rights reserved). Language: en

https://psycnet.apa.org/journals/ccp/84/6/544.pdf

A brief mobile app reduces nonsuicidal and suicidal self-injury: Evidence from three randomized controlled trials.

Behavioral and electrophysiological correlates of training-induced cognitive control improvements.

OBJECTIVE Mental illness is a leading cause of disease burden; however, many barriers prevent people from seeking mental health services. Technological innovations may improve our ability to reach underserved populations by overcoming many existing barriers. We evaluated a brief, automated risk assessment and intervention platform designed to increase the use of crisis resources provided to those online and in crisis. METHOD Participants, users of the digital mental health app Koko, were randomly assigned to treatment or control conditions upon accessing the app and were included in the study after their posts were identified by machine learning classifiers as signaling a current mental health crisis. Participants in the treatment condition received a brief Barrier Reduction Intervention (BRI) designed to increase the use of crisis service referrals provided on the app. Participants were followed up several hours later to assess the use of crisis services. RESULTS Only about one quarter of participants in a crisis (21.8%) reported being "very likely" to use clinical referrals provided to them, with the most commonly endorsed barriers being they "just want to chat" or their "thoughts are too intense." Among participants providing follow-up data (41.3%), receipt of the BRI was associated with a 23% increase in the use of crisis services. CONCLUSION These findings suggest that a brief, automated BRI can be efficacious on digital platforms, even among individuals experiencing acute psychological distress. The potential to increase help seeking and service utilization with such procedures holds promise for those in need of psychiatric services. TRIAL REGISTRATION clinicaltrials.gov identifier: NCT03633825. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Randomized controlled trial of an online machine learning-driven risk assessment and intervention platform for increasing the use of crisis services.

Researchers are increasingly using objective methods to study constructs related to suicidal thoughts and behavior (STB; e.g., self-identification with suicide), such as via behavioral tasks and brain imaging. Although promising, such approaches often are limited by the use of overly general stimuli (e.g., images, words) to represent constructs under study (e.g., suicide attempt). Overly general stimuli are problematic because they inadequately represent constructs, contributing to measurement error and thereby decreasing the internal and external validity of findings. To address this issue in suicide research specifically, we developed a set of first-person (FP) perspective suicide images depicting suicide attempt, evaluated its psychometric properties in one study, and examined whether people with recent suicidal thoughts rated FP-suicide images differently than people with no history of STB in a second study. Study 1 (N = 221) results suggest FP-suicide images have good construct validity (rs = .66-.87 with other suicide images) and internal consistency (Cronbach αs > .80) across three subjective rating dimensions (i.e., valence, arousal, threat). Study 2 (N = 73) results suggest that people with recent suicidal thoughts display substantially lower aversion toward FP-suicide images (i.e., lower negative valence [d = 1.22, p < .001], lower arousal [d = .61, p < .05] and lower threat ratings [d = 1.27, p < .001]) than people with no STB history. FP-suicide images provide researchers using behavioral tasks/brain imaging paradigms with more self-relevant stimuli that may increase the internal and external validity of findings. First-person stimuli may improve both our understanding and prediction of STB as well as provide novel targets (e.g., lower aversion to suicide) for clinical intervention. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Adam C. Jaroszewski

Papers

Risk factors for suicidal thoughts and behaviors: A meta-analysis of 50 years of research.

A brief mobile app reduces nonsuicidal and suicidal self-injury: Evidence from three randomized controlled trials.

Behavioral and electrophysiological correlates of training-induced cognitive control improvements.

Randomized controlled trial of an online machine learning-driven risk assessment and intervention platform for increasing the use of crisis services.

First-person stimuli: Improving the validity of stimuli in studies of suicide and related behaviors