Digital phenotyping of student mental health during COVID-19: an observational study of 100 college students.

Objective: This study assessed the feasibility of capturing smartphone based digital phenotyping data in college students during the COVID-19 pandemic with the goal of understanding how digital biomarkers of behavior correlate with mental health. Participants: Participants were 100 students enrolled in 4-year universities. Methods: Each participant attended a virtual visit to complete a series of gold-standard mental health assessments, and then used a mobile app for 28 days to complete mood assessments and allow for passive collection of GPS, accelerometer, phone call, and screen time data. Students completed another virtual visit at the end of the study to collect a second round of mental health assessments. Results: In-app daily mood assessments were strongly correlated with their corresponding gold standard clinical assessment. Sleep variance among students was correlated to depression scores (ρ = .28) and stress scores (ρ = .27). Conclusions: Digital Phenotyping among college students is feasible on both an individual and a sample level. Studies with larger sample sizes are necessary to understand population trends, but there are practical applications of the data today.

Challenges in initiating a lung cancer screening program: Experiences from two VA medical centers.

Establishing a lung cancer screening (LCS) program is an important endeavor that delivers life-saving healthcare to an at-risk population. However, developing a comprehensive LCS program requires critical elements including obtaining institutional level buy-in, hiring necessary personnel, developing appropriate infrastructure and actively engaging primary care providers, subspecialty services, and radiology. The process required to connect such services to deliver an organized LCS program that reaches all eligible candidates must be individualized to each institution's needs and infrastructure. Here we provide detailed experiences from two successful LCS programs, one using a primary care provider-based service and the other using a consult-based service. In each case, we provide the pros and cons of each system. We propose that the decision to setup an ideal LCS program could include a hybrid design that combines aspects of each system.

Barriers to Surveillance for Hepatocellular Carcinoma in a Multicenter Cohort.

Importance: Hepatocellular carcinoma (HCC) surveillance is underused in clinical practice, which may be owing to patient and clinician barriers. Objective: To characterize HCC surveillance barriers and associations with clinical outcomes in a multicenter cohort of patients with cirrhosis. Design, Setting, and Participants: This retrospective, multicenter cohort study included 5 medical centers in the United States. Patients with cirrhosis and newly diagnosed HCC treated from 2014 to 2018 were included. Data were analyzed from June 2021 to February 2022. Exposure: Surveillance completion in the 36-month period prior to HCC diagnosis. Main Outcomes and Measures: Surveillance receipt was classified as semiannual, annual, or no surveillance. Multivariable logistic regression analysis was used to identify factors associated with semiannual surveillance. We conducted multivariable logistic and Cox regression analyses to characterize associations between surveillance completion with curative treatment and overall survival. Results: A total 629 eligible patients (median [IQR] age, 63.6 [56.2-71.0] years; 491 [78.1%] men) were assessed, including 7 American Indian or Alaska Native patients (1.1%), 14 Asian patients (2.2), 176 Black patients (28.0%), 86 Hispanic patients (13.1%), and 340 White patients (54.1%). Nearly two-thirds of the cohort had no surveillance prior to HCC diagnosis (mean [range by site] 63.7% [37.9%-80.4%]), with a mean (range by site) of 14.0% (5.3%-33.3%) of patients having received semiannual surveillance and 22.3% (14.3%-28.8%) of patients having received annual surveillance. The most common reasons for no surveillance were lack of surveillance orders or nonadherence (mean [range by site], 82.4% [66.7%-92.4%], although a mean (range by site) of 17.6% (10.2%-22.1%) of patients had unrecognized cirrhosis at HCC presentation. Semiannual surveillance was associated with hepatitis B infection (odds ratio [OR], 3.06 [95% CI, 1.24-7.23]) and inversely associated with Black race (OR, 0.41 [95% CI, 0.20-0.80]) and lack of cirrhosis recognition (OR, 0.14 [95% CI, 0.02-0.46]). Semiannual HCC surveillance was significantly associated with curative treatment receipt (OR, 2.73 [95% CI, 1.60-4.70]) but not overall survival (HR, 0.81 [95% CI, 0.55-1.18]). Conclusions and Relevance: In this cohort study of patients with cirrhosis, HCC surveillance was underused in more than 80% of patients and associated with failures across the screening process. Dedicated programs to improve cirrhosis detection and HCC surveillance attainment are needed.

Similarity matrix-based anomaly detection for clinical intervention.

The use of digital phenotyping methods in clinical care has allowed for improved investigation of spatiotemporal behaviors of patients. Moreover, detecting abnormalities in mobile sensor data patterns can be instrumental in identifying potential changes in symptomology. We propose a method that temporally aligns sensor data in order to achieve interpretable measures of similarity between time points. These computed measures can then be used for anomaly detection, baseline routine computation, and trajectory clustering. In addition, we apply this method on a study of 695 college participants, as well as on a patient with worsening anxiety and depression. With varying temporal constraints, we find mild correlations between changes in routine and clinical scores. Furthermore, in our experiment on an individual with elevated depression and anxiety, we are able to cluster GPS trajectories, allowing for improved understanding and visualization of routines with respect to symptomology. In the future, we aim to apply this method on individuals that undergo data collection for longer periods of time, thus allowing for a better understanding of long-term routines and signals for clinical intervention.

Longitudinal symptom changes and association with home time in people with schizophrenia: An observational digital phenotyping study.

BACKGROUND: Smartphone assessments and sensors offer the ability to easily assess symptoms across environments in a naturalistic and longitudinal manner. However, the value of this new data to make inferences about personal vs population health and the role of environment in moderating symptoms in schizophrenia has not been fully explored in a scalable and reproducible manner. METHODS: Eighty-six adults with a diagnosis of schizophrenia were recruited from the Greater Boston Area between August 2019 and May 2021. Using the open-source mindLAMP app in an observational manner, smartphone surveys and sensors (GPS, accelerometer, screen on/off and call and text logs) were collected for up to six months. RESULTS: Sixty-three participants were analyzed, who had at least completed one survey in the app. App-based self-reported symptom surveys were highly correlated with scores on gold standard clinical assessments (r = 0.80, p = 10-11 for mood and r = 0.78, p = 10-12 for anxiety). For these app-based assessments, inter-individual differences account for a larger proportion of the correlations in longitudinal symptoms as compared to intra-individual differences. Mood, sleep, and psychosis symptoms reported on app surveys were more severe when taken at home as determined by the smartphone's GPS sensor. DISCUSSION: The intra-individual symptom correlations and the stratification of symptoms by home-time highlight the utility of digital phenotyping methods as a diagnostic tool, as well as the potential for personalized psychiatric treatment building on this data.

Prognosis conversations in advanced liver disease: A qualitative interview study with health professionals and patients.

Advanced Liver Disease (AdvLD) is common, morbid, and associated with high likelihood of death. Patients may not fully understand their prognosis and are often unprepared for the course of illness. Little is known about how and when to deliver prognosis-related information to patients with AdvLD, who should participate, and what should be discussed. We conducted in-depth interviews with a multi-profession sample of Hepatology clinicians and patients with AdvLD. Participants were drawn from three geographically diverse facilities (New England, Texas, California). We used inductive and deductive qualitative data analysis approaches to identify themes related to AdvLD prognosis discussions. Thematic analysis focused on content, timing, and participants' roles in prognosis discussions. In total, 31 patients with AdvLD and 26 multi-profession clinicians completed interviews. Most participants provided a broad conceptualization of prognosis beyond predictions of survival, including expectations about illness course, ways to manage or avoid complications and a need to address patients' emotions. Patients favored initiating discussions early in the AdvLD course and welcomed a multi-profession approach to conducting discussions. Clinicians favored a larger role for specialty physicians. All participants recognized that AdvLD prognosis discussions occur infrequently and favored a structured, standardized approach to broadly discussing prognosis. Patients with AdvLD and their clinicians favored a multifaceted approach to prognosis conversations including discussions of life expectancy, predictions about likely course of liver disease, and expected changes in function and capabilities over time. Structured and early prognosis discussions should be part of routine AdvLD care.

College student engagement with mental health apps: analysis of barriers to sustained use.

Objective: College students'demand for mental health resources is straining the services offered by colleges. While mobile apps demonstrate potential to help, students' engagement with these apps remains low. This study examines why college students show poor engagement with mental health apps and how apps may be adapted to suit this population. Participants: Participants were a convenience sample of 100 college students. Methods: Qualitative data was gathered through individual online interviews concerningattitudes toward mental health apps, and quantitative data was gathered through a survey about phone and app use. Results: Students were interested in mental health apps. 53% haddownloaded an app at one point, but only 19% currently used a mental health app. Stress and costdrove mental health app choices. Responses around engagement centered on: Data privacy, user interface, credibility, and customization. Conclusions: Students have specific wants for mental health apps including safety, simplicity, credibility, and customizability.

Two mental models of integrated care for advanced liver disease: qualitative study of multidisciplinary health professionals.

OBJECTIVES: The purpose of this paper is to present two divergent mental models of integrated advanced liver disease (AdvLD) care among 26 providers who treat patients with AdvLD. SETTING: 3 geographically dispersed United States Veterans Health Administration health systems. PARTICIPANTS: 26 professionals (20 women and 6 men) participated, including 9 (34.6%) gastroenterology, hepatology, and transplant physicians, 2 (7.7%) physician assistants, 7 (27%) nurses and nurse practitioners, 3 (11.5%) social workers and psychologists, 4 (15.4%) palliative care providers and 1 (3.8%) pharmacist. MAIN OUTCOME MEASURES: We conducted qualitative in-depth interviews of providers caring for patients with AdvLD. We used framework analysis to identify two divergent mental models of integrated AdvLD care. These models vary in timing of initiating various constituents of care, philosophy of integration, and supports and resources needed to achieve each model. RESULTS: Clinicians described integrated care as an approach that incorporates elements of curative care, symptom and supportive care, advance care planning and end-of-life services from a multidisciplinary team. Analysis revealed two mental models that varied in how and when these constituents are delivered. One mental model involves sequential transitions between constituents of care, and the second mental model involves synchronous application of the various constituents. Participants described elements of teamwork and coordination supports necessary to achieve integrated AdvLD care. Many discussed the importance of having a multidisciplinary team integrating supportive care, symptom management and palliative care with liver disease care. CONCLUSIONS: Health professionals agree on the constituents of integrated AdvLD care but describe two competing mental models of how these constituents are integrated. Health systems can promote integrated care by assembling multidisciplinary teams, and providing teamwork and coordination supports, and training that facilitates patient-centred AdvLD care.

Applying machine learning to smartphone based cognitive and sleep assessments in schizophrenia.

BACKGROUND: Cognitive impairment in schizophrenia remains a chief source of functional disability and impairment, despite the potential for effective interventions. This is in part related to a lack of practical and easy to administer screening strategies that can identify and help triage cognitive impairment. This study explores how smartphone-based assessments may help address this need. METHODS: In this study, data was analyzed from 25 subjects with schizophrenia and 30 controls who engaged with a gamified mobile phone version of the Trails-B cognitive assessment in their everyday life over 90 days and complete a clinical neurocognitive testing battery at the beginning and end of the study. Machine learning was applied to the resulting dataset to predict disease status and neurocognitive function and understand which features were most important for accurate prediction. RESULTS: The generated models predicted disease status with high accuracy using static features alone (AUC = 0.94), with the total number of items collected and the total duration of interaction with the application most predictive. The addition of temporal data statistically significantly improved performance (AUC = 0.95), with the amount of idle time a significant new predictor. Correlates of sleep dysfunction were also predicted (AUC = 0.80), with similar feature importance. DISCUSSION: Machine learning enabled the highly accurate identification of subjects with schizophrenia versus healthy controls, and the accurate prediction of neurocognitive function. The addition of temporal data significantly improved the performance of these models, underscoring the value of smartphone-based assessments of cognition as a practical tool for assessing cognition.

Digital phenotyping for mental health of college students: a clinical review.

Experiencing continued growth in demand for mental health services among students, colleges are seeking digital solutions to increase access to care as classes shift to remote virtual learning during the COVID-19 pandemic. Using smartphones to capture real-time symptoms and behaviours related to mental illnesses, digital phenotyping offers a practical tool to help colleges remotely monitor and assess mental health and provide more customised and responsive care. This narrative review of 25 digital phenotyping studies with college students explored how this method has been deployed, studied and has impacted mental health outcomes. We found the average duration of studies to be 42 days and the average enrolled to be 81 participants. The most common sensor-based streams collected included location, accelerometer and social information and these were used to inform behaviours such as sleep, exercise and social interactions. 52% of the studies included also collected smartphone survey in some form and these were used to assess mood, anxiety and stress among many other outcomes. The collective focus on data that construct features related to sleep, activity and social interactions indicate that this field is already appropriately attentive to the primary drivers of mental health problems among college students. While the heterogeneity of the methods of these studies presents no reliable target for mobile devices to offer automated help-the feasibility across studies suggests the potential to use these data today towards personalising care. As more unified digital phenotyping research evolves and scales to larger sample sizes, student mental health centres may consider integrating these data into their clinical practice for college students.

Smartphone Apps for College Mental Health: A Concern for Privacy and Quality of Current Offerings.

OBJECTIVE: The expanding need for mental health services on college campuses has increased interest in smartphone apps for mental health. Although apps can be effective tools, they can also present risks regarding privacy and lack of efficacy, and little is known about the nature or quality of mental health apps that colleges are recommending to students. This study sought to fill that knowledge gap. METHODS: Sixty college counseling center Web sites were examined for suggested mobile apps offered as resources for students. The features of each app were assessed, including date of last update, privacy policy, and whether any research had been published in peer-reviewed journals. RESULTS: Twenty-six college counseling centers suggested a total of 218 unique apps. Of these, 28% were no longer available for download. Of the 158 remaining apps, only 44% had been updated in the past 6 months, and 39% had no privacy policy. Of the 97 existing apps with privacy policies, 88% collected user's data and 49% shared users' data with third parties. Efficacy studies had been published in peer-reviewed journals for only 16% of existing apps. CONCLUSIONS: College counseling centers are not suggesting safe or up-to-date mobile apps. There is an urgent need to help centers identify, curate, and recommend more appropriate apps. The authors suggest that centers use the American Psychiatric Association's app evaluation framework; develop an app review process, with input from clinicians and students; or employ a digital navigator to select apps and help connect students to the appropriate app.

Elevated Acoustic Startle Responses in Humans: Relationship to Reduced Loudness Discomfort Level, but not Self-Report of Hyperacusis.

Increases in the acoustic startle response (ASR) of animals have been reported following experimental manipulations to induce tinnitus, an auditory disorder defined by phantom perception of sound. The increases in ASR have been proposed to signify the development of hyperacusis, a clinical condition defined by intolerance of normally tolerable sound levels. To test this proposal, the present study compared ASR amplitude to measures of sound-level tolerance (SLT) in humans, the only species in which SLT can be directly assessed. Participants had clinically normal/near-normal hearing thresholds, were free of psychotropic medications, and comprised people with tinnitus and without. ASR was measured as eyeblink-related electromyographic activity in response to a noise pulse presented at a range of levels and in two background conditions (noise and quiet). SLT was measured as loudness discomfort level (LDL), the lowest level of sound deemed uncomfortable, and via a questionnaire on the loudness of sounds in everyday life. Regardless of tinnitus status, ASR amplitude at a given stimulus level increased with decreasing LDL, but showed no relationship to SLT self-reported via the questionnaire. These relationships (or lack thereof) could not be attributed to hearing threshold, age, anxiety, or depression. The results imply that increases in ASR in the animal work signify decreases in LDL specifically and may not correspond to the development of hyperacusis as would be self-reported by a clinic patient.

Increased contralateral suppression of otoacoustic emissions indicates a hyperresponsive medial olivocochlear system in humans with tinnitus and hyperacusis.

Atypical medial olivocochlear (MOC) feedback from brain stem to cochlea has been proposed to play a role in tinnitus, but even well-constructed tests of this idea have yielded inconsistent results. In the present study, it was hypothesized that low sound tolerance (mild to moderate hyperacusis), which can accompany tinnitus or occur on its own, might contribute to the inconsistency. Sound-level tolerance (SLT) was assessed in subjects (all men) with clinically normal or near-normal thresholds to form threshold-, age-, and sex-matched groups: 1) no tinnitus/high SLT, 2) no tinnitus/low SLT, 3) tinnitus/high SLT, and 4) tinnitus/low SLT. MOC function was measured from the ear canal as the change in magnitude of distortion-product otoacoustic emissions (DPOAE) elicited by broadband noise presented to the contralateral ear. The noise reduced DPOAE magnitude in all groups ("contralateral suppression"), but significantly more reduction occurred in groups with tinnitus and/or low SLT, indicating hyperresponsiveness of the MOC system compared with the group with no tinnitus/high SLT. The results suggest hyperresponsiveness of the interneurons of the MOC system residing in the cochlear nucleus and/or MOC neurons themselves. The present data, combined with previous human and animal data, indicate that neural pathways involving every major division of the cochlear nucleus manifest hyperactivity and/or hyperresponsiveness in tinnitus and/or low SLT. The overactivation may develop in each pathway separately. However, a more parsimonious hypothesis is that top-down neuromodulation is the driving force behind ubiquitous overactivation of the auditory brain stem and may correspond to attentional spotlighting on the auditory domain in tinnitus and hyperacusis.

Subcallosal brain structure: correlation with hearing threshold at supra-clinical frequencies (>8 kHz), but not with tinnitus.

This study tested for differences in brain structure between tinnitus and control subjects, focusing on a subcallosal brain region where striking differences have been inconsistently found previously. Voxel-based morphometry (VBM) was used to compare structural MRIs of tinnitus subjects and non-tinnitus controls. Audiograms of all subjects were normal or near-normal at standard clinical frequencies (≤8 kHz). Mean threshold through 14 kHz, age, sex and handedness were matched between groups. There were no definitive differences between tinnitus and control groups in modulated or unmodulated maps of gray matter (GM) probability (i.e., GM volume and concentration, respectively). However, when the image data were tested for correlations with parameters that were either not measured or not matched between the tinnitus and control groups of previous studies, a notable correlation was found: Threshold at supra-clinical frequencies (above 8 kHz) was negatively correlated with modulated GM probability in ventral posterior cingulate cortex, dorsomedial prefrontal cortex, and a subcallosal region that included ventromedial prefrontal cortex and coincided with previously-reported differences between tinnitus and control subjects. The results suggest an explanation for the discrepant findings in subcallosal brain: threshold at supra-clinical frequencies may have differed systematically between tinnitus and control groups in some studies but not others. The observed correlation between (1) brain structure in regions engaged in cognitive and attentional processes and (2) hearing sensitivity at frequencies generally considered unnecessary for normal human auditory behavior is surprising and worthy of follow up.

Brainstem auditory evoked potentials suggest a role for the ventral cochlear nucleus in tinnitus.

Numerous studies have demonstrated elevated spontaneous and sound-evoked brainstem activity in animal models of tinnitus, but data on brainstem function in people with this common clinical condition are sparse. Here, auditory nerve and brainstem function in response to sound was assessed via auditory brainstem responses (ABR) in humans with tinnitus and without. Tinnitus subjects showed reduced wave I amplitude (indicating reduced auditory nerve activity) but enhanced wave V (reflecting elevated input to the inferior colliculi) compared with non-tinnitus subjects matched in age, sex, and pure-tone threshold. The transformation from reduced peripheral activity to central hyperactivity in the tinnitus group was especially apparent in the V/I and III/I amplitude ratios. Compared with a third cohort of younger, non-tinnitus subjects, both tinnitus, and matched, non-tinnitus groups showed elevated thresholds above 4 kHz and reduced wave I amplitude, indicating that the differences between tinnitus and matched non-tinnitus subjects occurred against a backdrop of shared peripheral dysfunction that, while not tinnitus specific, cannot be discounted as a factor in tinnitus development. Animal lesion and human neuroanatomical data combine to indicate that waves III and V in humans reflect activity in a pathway originating in the ventral cochlear nucleus (VCN) and with spherical bushy cells (SBC) in particular. We conclude that the elevated III/I and V/I amplitude ratios in tinnitus subjects reflect disproportionately high activity in the SBC pathway for a given amount of peripheral input. The results imply a role for the VCN in tinnitus and suggest the SBC pathway as a target for tinnitus treatment.

Hearing without listening: functional connectivity reveals the engagement of multiple nonauditory networks during basic sound processing.

The present functional magnetic resonance imaging (fMRI) study presents data challenging the traditional view that sound is processed almost exclusively in the classical auditory pathway unless imbued with behavioral significance. In a first experiment, subjects were presented with broadband noise in on/off fashion as they performed an unrelated visual task. A conventional analysis assuming predictable sound-evoked responses demonstrated a typical activation pattern that was confined to classical auditory centers. In contrast, spatial independent component analysis (sICA) disclosed multiple networks of acoustically responsive brain centers. One network comprised classical auditory centers, but four others included nominally "nonauditory" areas: cingulo-insular cortex, mediotemporal limbic lobe, basal ganglia, and posterior orbitofrontal cortex, respectively. Functional connectivity analyses confirmed the sICA results by demonstrating coordinated activity between the involved brain structures. In a second experiment, fMRI data obtained from unstimulated (i.e., resting) subjects revealed largely similar networks. Together, these two experiments suggest the existence of a coordinated system of multiple acoustically responsive intrinsic brain networks, comprising classical auditory centers but also other brain areas. Our results suggest that nonauditory centers play a role in sound processing at a very basic level, even when the sound is not intertwined with behaviors requiring the well-known functionality of these regions.

Ringing ears: the neuroscience of tinnitus.

Tinnitus is a phantom sound (ringing of the ears) that affects quality of life for millions around the world and is associated in most cases with hearing impairment. This symposium will consider evidence that deafferentation of tonotopically organized central auditory structures leads to increased neuron spontaneous firing rates and neural synchrony in the hearing loss region. This region covers the frequency spectrum of tinnitus sounds, which are optimally suppressed following exposure to band-limited noise covering the same frequencies. Cross-modal compensations in subcortical structures may contribute to tinnitus and its modulation by jaw-clenching and eye movements. Yet many older individuals with impaired hearing do not have tinnitus, possibly because age-related changes in inhibitory circuits are better preserved. A brain network involving limbic and other nonauditory regions is active in tinnitus and may be driven when spectrotemporal information conveyed by the damaged ear does not match that predicted by central auditory processing.

Tinnitus, diminished sound-level tolerance, and elevated auditory activity in humans with clinically normal hearing sensitivity.

Phantom sensations and sensory hypersensitivity are disordered perceptions that characterize a variety of intractable conditions involving the somatosensory, visual, and auditory modalities. We report physiological correlates of two perceptual abnormalities in the auditory domain: tinnitus, the phantom perception of sound, and hyperacusis, a decreased tolerance of sound based on loudness. Here, subjects with and without tinnitus, all with clinically normal hearing thresholds, underwent 1) behavioral testing to assess sound-level tolerance and 2) functional MRI to measure sound-evoked activation of central auditory centers. Despite receiving identical sound stimulation levels, subjects with diminished sound-level tolerance (i.e., hyperacusis) showed elevated activation in the auditory midbrain, thalamus, and primary auditory cortex compared with subjects with normal tolerance. Primary auditory cortex, but not subcortical centers, showed elevated activation specifically related to tinnitus. The results directly link hyperacusis and tinnitus to hyperactivity within the central auditory system. We hypothesize that the tinnitus-related elevations in cortical activation may reflect undue attention drawn to the auditory domain, an interpretation consistent with the lack of tinnitus-related effects subcortically where activation is less potently modulated by attentional state. The data strengthen, at a mechanistic level, analogies drawn previously between tinnitus/hyperacusis and other, nonauditory disordered perceptions thought to arise from neural hyperactivity such as chronic neuropathic pain and photophobia.