Evaluation of Commercially Available Machine Interpretation Applications for Simple Clinical Communication.

BACKGROUND: Accessing professional medical interpreters for brief, low risk exchanges can be challenging. Machine translation (MT) for verbal communication has the potential to be a useful clinical tool, but few evaluations exist. OBJECTIVE: We evaluated the quality of three MT applications for English-Spanish and English-Mandarin two-way interpretation of low complexity brief clinical communication compared with human interpretation. DESIGN: Audio-taped phrases were interpreted via human and 3 MT applications. Bilingual assessors evaluated the quality of MT interpretation on four assessment categories (accuracy, fluency, meaning, and clinical risk) using 5-point Likert scales. We used a non-inferiority design with 15% inferiority margin to evaluate the quality of three MT applications with professional medical interpreters serving as gold standards. MAIN MEASURES: Proportion of interpretation exchanges deemed acceptable, defined as a composite score of 16 or greater out of 20 based on the four assessment categories. KEY RESULTS: For English to Spanish, the proportion of MT-interpreted phrases scored as acceptable ranged from 0.68 to 0.84, while for English to Mandarin, the range was from 0.62 to 0.76. Both Spanish/Mandarin to English MT interpretation had low acceptable scores (range 0.36 to 0.41). No MT interpretation met the non-inferiority threshold. CONCLUSION: While MT interpretation was better for English to Spanish or Mandarin than the reverse, the overall quality of MT interpretation was poor for two-way clinical communication. Clinicians should advocate for easier access to professional interpretation in all clinical spaces and defer use of MT until these applications improve.

Satisfaction can co-exist with hesitation: qualitative analysis of acceptability of telemedicine among multi-lingual patients in a safety-net healthcare system during the COVID-19 pandemic.

BACKGROUND: The COVID-19 pandemic triggered unprecedented expansion of outpatient telemedicine in the United States in all types of health systems, including safety-net health systems. These systems generally serve low-income, racially/ethnically/linguistically diverse patients, many of whom face barriers to digital health access. These patients' perspectives are vital to inform ongoing, equitable implementation efforts. METHODS: Twenty-five semi-structured interviews exploring a theoretical framework of technology acceptability were conducted from March through July 2020. Participants had preferred languages of English, Spanish, or Cantonese and were recruited from three clinics (general medicine, obstetrics, and pulmonary) within the San Francisco Health Network. Both deductive and inductive coding were performed. In a secondary analysis, qualitative data were merged with survey data to relate perspectives to demographic factors and technology access/use. RESULTS: Participants were diverse with respect to language (52% non-English-speaking), age (range 23-71), race/ethnicity (24% Asian, 20% Black, 44% Hispanic/Latinx, 12% White), & smartphone use (80% daily, 20% weekly or less). All but 2 had a recent telemedicine visit (83% telephone). Qualitative results revealed that most participants felt telemedicine visits fulfilled their medical needs, were convenient, and were satisfied with their telemedicine care. However, most still preferred in-person visits, expressing concern that tele-visits relied on patients' abilities to access telemedicine, as well as monitor and manage their own health without in-person physical evaluation. CONCLUSIONS: High satisfaction with telemedicine can co-exist with patient-expressed hesitations surrounding the perceived effectiveness, self-efficacy, and digital access barriers associated with a new model of care. More research is needed to guide how healthcare systems and clinicians make decisions and communicate about visit modalities to support high-quality care that responds to patients' needs and circumstances.

The Impact of the COVID-19 Pandemic on Internet Use and the Use of Digital Health Tools: Secondary Analysis of the 2020 Health Information National Trends Survey.

BACKGROUND: The COVID-19 pandemic increased the use of digital tools in health care (eg, patient portal, telemedicine, and web-based scheduling). Studies have shown that older individuals, racial/ethnic minority groups, or populations with lower educational attainment or income have lower rates of using digital health tools. Digitalization of health care may exacerbate already existing access barriers in these populations. OBJECTIVE: This study evaluated how use of digital tools to asynchronously communicate with clinicians, schedule appointments, and view medical records changed near the beginning of the pandemic. METHODS: Using 2020 Health Information National Trends Survey (HINTS) data, we examined internet use and 7 digital health technology use outcomes (electronic communication with a provider, electronic appointment scheduling, electronic test result viewing, patient portal access, portal use to download health records, portal use for patient-provider communication, and portal use to view test results). The HINTS surveyors designated surveys received after March 11, 2020, as postpandemic responses. Using weighted logistic regression, we investigated the impact of the pandemic after adjusting for sociodemographic traits (age, race/ethnicity, income, education, and gender), digital access (having ever used the internet and smartphone/tablet ownership), and health-related factors (insurance coverage, caregiver status, having a regular provider, and chronic diseases). To explore differences in changes in outcomes among key sociodemographic groups, we tested for significant interaction terms between the pandemic variable and race/ethnicity, age, income, and educational attainment. RESULTS: There were 3865 respondents (1437 prepandemic and 2428 postpandemic). Of the 8 outcomes investigated, the pandemic was only significantly associated with higher odds (adjusted odds ratio 1.99, 95% CI 1.18-3.35) of using electronic communication with a provider. There were significant interactions between the pandemic variable and 2 key sociodemographic traits. Relative to the lowest income group (

Hippo Cascade Controls Lineage Commitment of Liver Tumors in Mice and Humans.

Primary liver cancer consists mainly of hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC). A subset of human HCCs expresses a ICC-like gene signature and is classified as ICC-like HCC. The Hippo pathway is a critical regulator of normal and malignant liver development. However, the precise function(s) of the Hippo cascade along liver carcinogenesis remain to be fully delineated. The role of the Hippo pathway in a murine mixed HCC/ICC model induced by activated forms of AKT and Ras oncogenes (AKT/Ras) was investigated. The authors demonstrated the inactivation of Hippo in AKT/Ras liver tumors leading to nuclear localization of Yap and TAZ. Coexpression of AKT/Ras with Lats2, which activates Hippo, or the dominant negative form of TEAD2 (dnTEAD2), which blocks Yap/TAZ activity, resulted in delayed hepatocarcinogenesis and elimination of ICC-like lesions in the liver. Mechanistically, Notch2 expression was found to be down-regulated by the Hippo pathway in liver tumors. Overexpression of Lats2 or dnTEAD2 in human HCC cell lines inhibited their growth and led to the decreased expression of ICC-like markers, as well as Notch2 expression. Altogether, this study supports the key role of the Hippo cascade in regulating the differentiation status of liver tumors.

Leveraging electronic health records and knowledge networks for Alzheimer's disease prediction and sex-specific biological insights.

Identification of Alzheimer's disease (AD) onset risk can facilitate interventions before irreversible disease progression. We demonstrate that electronic health records from the University of California, San Francisco, followed by knowledge networks (for example, SPOKE) allow for (1) prediction of AD onset and (2) prioritization of biological hypotheses, and (3) contextualization of sex dimorphism. We trained random forest models and predicted AD onset on a cohort of 749 individuals with AD and 250,545 controls with a mean area under the receiver operating characteristic of 0.72 (7 years prior) to 0.81 (1 day prior). We further harnessed matched cohort models to identify conditions with predictive power before AD onset. Knowledge networks highlight shared genes between multiple top predictors and AD (for example, APOE, ACTB, IL6 and INS). Genetic colocalization analysis supports AD association with hyperlipidemia at the APOE locus, as well as a stronger female AD association with osteoporosis at a locus near MS4A6A. We therefore show how clinical data can be utilized for early AD prediction and identification of personalized biological hypotheses.

Standardized Integration of Person-Generated Data Into Routine Clinical Care.

Person-generated data (PGD) are a valuable source of information on a person's health state in daily life and in between clinic visits. To fully extract value from PGD, health care organizations must be able to smoothly integrate data from PGD devices into routine clinical workflows. Ideally, to enhance efficiency and flexibility, such integrations should follow reusable processes that can easily be replicated for multiple devices and data types. Instead, current PGD integrations tend to be one-off efforts entailing high costs to build and maintain custom connections with each device and their proprietary data formats. This viewpoint paper formulates the integration of PGD into clinical systems and workflow as a PGD integration pipeline and reviews the functional components of such a pipeline. A PGD integration pipeline includes PGD acquisition, aggregation, and consumption. Acquisition is the person-facing component that includes both technical (eg, sensors, smartphone apps) and policy components (eg, informed consent). Aggregation pools, standardizes, and structures data into formats that can be used in health care settings such as within electronic health record-based workflows. PGD consumption is wide-ranging, by different solutions in different care settings (inpatient, outpatient, consumer health) for different types of users (clinicians, patients). The adoption of data and metadata standards, such as those from IEEE and Open mHealth, would facilitate aggregation and enable broader consumption. We illustrate the benefits of a standards-based integration pipeline for the illustrative use case of home blood pressure monitoring. A standards-based PGD integration pipeline can flexibly streamline the clinical use of PGD while accommodating the complexity, scale, and rapid evolution of today's health care systems.

Deep phenotyping of Alzheimer's disease leveraging electronic medical records identifies sex-specific clinical associations.

Alzheimer's Disease (AD) is a neurodegenerative disorder that is still not fully understood. Sex modifies AD vulnerability, but the reasons for this are largely unknown. We utilize two independent electronic medical record (EMR) systems across 44,288 patients to perform deep clinical phenotyping and network analysis to gain insight into clinical characteristics and sex-specific clinical associations in AD. Embeddings and network representation of patient diagnoses demonstrate greater comorbidity interactions in AD in comparison to matched controls. Enrichment analysis identifies multiple known and new diagnostic, medication, and lab result associations across the whole cohort and in a sex-stratified analysis. With this data-driven method of phenotyping, we can represent AD complexity and generate hypotheses of clinical factors that can be followed-up for further diagnostic and predictive analyses, mechanistic understanding, or drug repurposing and therapeutic approaches.

Reversal of cancer gene expression identifies repurposed drugs for diffuse intrinsic pontine glioma.

Diffuse intrinsic pontine glioma (DIPG) is an aggressive incurable brainstem tumor that targets young children. Complete resection is not possible, and chemotherapy and radiotherapy are currently only palliative. This study aimed to identify potential therapeutic agents using a computational pipeline to perform an in silico screen for novel drugs. We then tested the identified drugs against a panel of patient-derived DIPG cell lines. Using a systematic computational approach with publicly available databases of gene signature in DIPG patients and cancer cell lines treated with a library of clinically available drugs, we identified drug hits with the ability to reverse a DIPG gene signature to one that matches normal tissue background. The biological and molecular effects of drug treatment was analyzed by cell viability assay and RNA sequence. In vivo DIPG mouse model survival studies were also conducted. As a result, two of three identified drugs showed potency against the DIPG cell lines Triptolide and mycophenolate mofetil (MMF) demonstrated significant inhibition of cell viability in DIPG cell lines. Guanosine rescued reduced cell viability induced by MMF. In vivo, MMF treatment significantly inhibited tumor growth in subcutaneous xenograft mice models. In conclusion, we identified clinically available drugs with the ability to reverse DIPG gene signatures and anti-DIPG activity in vitro and in vivo. This novel approach can repurpose drugs and significantly decrease the cost and time normally required in drug discovery.

OCTAD: an open workspace for virtually screening therapeutics targeting precise cancer patient groups using gene expression features.

As the field of precision medicine progresses, treatments for patients with cancer are starting to be tailored to their molecular as well as their clinical features. The emerging cancer subtypes defined by these molecular features require that dedicated resources be used to assist the discovery of drug candidates for preclinical evaluation. Voluminous gene expression profiles of patients with cancer have been accumulated in public databases, enabling the creation of cancer-specific expression signatures. Meanwhile, large-scale gene expression profiles of cellular responses to chemical compounds have also recently became available. By matching the cancer-specific expression signature to compound-induced gene expression profiles from large drug libraries, researchers can prioritize small molecules that present high potency to reverse expression of signature genes for further experimental testing of their efficacy. This approach has proven to be an efficient and cost-effective way to identify efficacious drug candidates. However, the success of this approach requires multiscale procedures, imposing considerable challenges to many labs. To address this, we developed Open Cancer TherApeutic Discovery (OCTAD; http://octad.org ): an open workspace for virtually screening compounds targeting precise groups of patients with cancer using gene expression features. Its database includes 19,127 patient tissue samples covering more than 50 cancer types and expression profiles for 12,442 distinct compounds. The program is used to perform deep-learning-based reference tissue selection, disease gene expression signature creation, drug reversal potency scoring and in silico validation. OCTAD is available as a web portal and a standalone R package to allow experimental and computational scientists to easily navigate the tool.