A multi-center validation of the electronic health record admission source and discharge location fields against the clinical notes for identifying inpatients with long-term care facility exposure

Identifying long-term care facility (LTCF)-exposed inpatients is important for infection control research and practice, but ascertaining LTCF exposure is challenging. Across a large validation study, electronic health record data fields identified 76% of LTCF-exposed patients compared to manual chart review.

OBJECTIVE

Residence or recent stay in a long-term care facility (LTCF) is an important risk factor for antibiotic-resistant bacterial colonization. However, absent dedicated intake questionnaires or resource-intensive chart review, ascertaining LTCF exposure in inpatients is challenging. We aimed to validate the electronic health record (EHR) admission and discharge location fields against the clinical notes for identifying LTCF-exposed inpatients.

METHODS

We conducted a retrospective study of 1020 randomly sampled adult admissions between 2016 and 2021 across 12 University of Maryland Medical System hospitals. Using study-developed guidelines, we categorized the following data for LTCF exposure: each admission’s history & physical (H&P) note, each admission’s EHR-extracted “Admission Source,” and (3) the EHR-extracted admission and discharge locations for previous admissions (≤90 days). We estimated sensitivities, with 95% CIs, of H&P notes and of EHR admission/discharge location fields for detecting “current” and “any recent” (≤90 days, including current) LTCF exposure.

RESULTS

For detecting current LTCF exposure, the sensitivity of the index admission’s EHR-extracted “Admission Source” was 46% (95% CI: 35%–58%) and of the H&P note was 92% (83%–97%). For detecting any recent LTCF exposure, the sensitivity of “Admission Source” across the index and previous admissions was 32% (24%–41%), “Discharge Location” across previous admission(s) was 57% (47%–66%), and of the H&P note was 68% (59%–76%). The combined sensitivity of admission source and discharge location for detecting any recent LTCF exposure was 76% (67%–83%).

CONCLUSIONS

The EHR-obtained admission source and discharge location fields identified 76% of LTCF-exposed patients compared to chart review but disproportionately missed currently exposed patients.

Neural Dynamics of the Processing of Speech Features: Evidence for a Progression of Features from Acoustic to Sentential Processing

When we listen to speech, our brain's neurophysiological responses "track" its acoustic features, but it is less well understood how these auditory responses are modulated by linguistic content. Here, we recorded magnetoencephalography (MEG) responses while subjects listened to four types of continuous-speech-like passages: speech-envelope modulated noise, English-like non-words, scrambled words, and narrative passage. Temporal response function (TRF) analysis provides strong neural evidence for the emergent features of speech processing in cortex, from acoustics to higher-level linguistics, as incremental steps in neural speech processing. Critically, we show a stepwise hierarchical progression of progressively higher order features over time, reflected in both bottom-up (early) and top-down (late) processing stages. Linguistically driven top-down mechanisms take the form of late N400-like responses, suggesting a central role of predictive coding mechanisms at multiple levels. As expected, the neural processing of lower-level acoustic feature responses is bilateral or right lateralized, with left lateralization emerging only for lexical-semantic features. Finally, our results identify potential neural markers of the computations underlying speech perception and comprehension.

Eelbrain, a Python toolkit for time-continuous analysis with temporal response functions

Even though human experience unfolds continuously in time, it is not strictly linear; instead, it entails cascading processes building hierarchical cognitive structures. For instance, during speech perception, humans transform a continuously varying acoustic signal into phonemes, words, and meaning, and these levels all have distinct but interdependent temporal structures. Time-lagged regression using temporal response functions (TRFs) has recently emerged as a promising tool for disentangling electrophysiological brain responses related to such complex models of perception. Here, we introduce the Eelbrain Python toolkit, which makes this kind of analysis easy and accessible. We demonstrate its use, using continuous speech as a sample paradigm, with a freely available EEG dataset of audiobook listening. A companion GitHub repository provides the complete source code for the analysis, from raw data to group-level statistics. More generally, we advocate a hypothesis-driven approach in which the experimenter specifies a hierarchy of time-continuous representations that are hypothesized to have contributed to brain responses, and uses those as predictor variables for the electrophysiological signal. This is analogous to a multiple regression problem, but with the addition of a time dimension. TRF analysis decomposes the brain signal into distinct responses associated with the different predictor variables by estimating a multivariate TRF (mTRF), quantifying the influence of each predictor on brain responses as a function of time(-lags). This allows asking two questions about the predictor variables: (1) Is there a significant neural representation corresponding to this predictor variable? And if so, (2) what are the temporal characteristics of the neural response associated with it? Thus, different predictor variables can be systematically combined and evaluated to jointly model neural processing at multiple hierarchical levels. We discuss applications of this approach, including the potential for linking algorithmic/representational theories at different cognitive levels to brain responses through computational models with appropriate linking hypotheses.

Donor activity is associated with US legislators' attention to political issues

Campaign contributions are a staple of congressional life. Yet, the search for tangible effects of congressional donations often focuses on the association between contributions and votes on congressional bills. We present an alternative approach by considering the relationship between money and legislators' speech. Floor speeches are an important component of congressional behavior, and reflect a legislator's policy priorities and positions in a way that voting cannot. Our research provides the first comprehensive analysis of the association between a legislator's campaign donors and the policy issues they prioritize with congressional speech. Ultimately, we find a robust relationship between donors and speech, indicating a more pervasive role of money in politics than previously assumed. We use a machine learning framework on a new dataset that brings together legislator metadata for all representatives in the US House between 1995 and 2018, including committee assignments, legislative speech, donation records, and information about Political Action Committees. We compare information about donations against other potential explanatory variables, such as party affiliation, home state, and committee assignments, and find that donors consistently have the strongest association with legislators' issue-attention. We further contribute a procedure for identifying speech and donation events that occur in close proximity to one another and share meaningful connections, identifying the proverbial needles in the haystack of speech and donation activity in Congress which may be cases of interest for investigative journalism. Taken together, our framework, data, and findings can help increase the transparency of the role of money in politics.

Evaluating the Use of Online Self-Report Questionnaires as Clinically Valid Mental Health Monitoring Tools in the Clinical Whitespace

Although digital health solutions are increasingly popular in clinical psychiatry, one application that has not been fully explored is the utilization of survey technology to monitor patients outside of the clinic. Supplementing routine care with digital information collected in the "clinical whitespace" between visits could improve care for patients with severe mental illness. This study evaluated the feasibility and validity of using online self-report questionnaires to supplement in-person clinical evaluations in persons with and without psychiatric diagnoses. We performed a rigorous in-person clinical diagnostic and assessment battery in 54 participants with schizophrenia (N = 23), depressive disorder (N = 14), and healthy controls (N = 17) using standard assessments for depressive and psychotic symptomatology. Participants were then asked to complete brief online assessments of depressive (Quick Inventory of Depressive Symptomatology) and psychotic (Community Assessment of Psychic Experiences) symptoms outside of the clinic for comparison with the ground-truth in-person assessments. We found that online self-report ratings of severity were significantly correlated with the clinical assessments for depression (two assessments used: R = 0.63, p < 0.001; R = 0.73, p < 0.001) and psychosis (R = 0.62, p < 0.001). Our results demonstrate the feasibility and validity of collecting psychiatric symptom ratings through online surveys. Surveillance of this kind may be especially useful in detecting acute mental health crises between patient visits and can generally contribute to more comprehensive psychiatric treatment.

1202. A Multi-Center Validation of the Electronic Health Record ‘Admission Source’ Field Against Clinical Notes for Identifying Hospitalized Patients with Long-term Care Facility Exposure

Background

A current or recent long-term care facility (LTCF) stay is a strong risk factor for antibiotic-resistant bacterial colonization and infection. However, most electronic health record (EHR) systems do not systematically record LTCF exposure. Absent manual chart review, which is resource-intensive and cannot be incorporated into automated screening algorithms, there is no definitive method for identifying LTCF-exposed inpatients. As a surrogate, researchers often use ‘Admission Source’ to identify LTCF transfers, but this EHR field has not been previously validated, and it may miss the unknown percentage of patients with recent, but not current, LTCF stays. This study evaluated the accuracy of ‘Admission Source’ in identifying LTCF-exposed inpatients.

Methods

This was a retrospective study of adult admissions from 2018 – 2021 across 12 hospitals in the University of Maryland Medical System. We extracted patient and encounter data and classified patients as LTCF-exposed by ‘Admission Source’ if they were LTCF transfers. For 315 randomly sampled admissions, M.T. and K.G. reviewed the admission ‘History & Physical’ note for mention of LTCF exposure (Fig. 1). Assuming an indication of LTCF in either the ‘Admission Source’ field or clinical note represented true exposure, we estimated each method’s sensitivity with 95% confidence intervals.

Results

Across 280,581 admissions, 9,476 (3.4%) had an ‘LTCF transfer’ admission source. In the validation sample, 26 (8.3%) were classified as LTCF-exposed by either ‘Admission Source’ or clinical note, of which 12 were identified in the ‘Admission Source’ field and 25 in the notes (Fig. 2). The sensitivity of ‘Admission Source’ for detecting LTCF exposure was 46% (29% – 65%) and for clinical notes was 96% (81% - 99%). Most (12/14) patients missed by ‘Admission Source’ were current LTCF residents (Fig. 3).

Conclusion

The EHR ‘Admission Source’ field misses the majority of inpatients with recent or current LTCF exposure, risking substantial misclassification in research studies and clinical algorithms that incorporate this variable. Automated techniques for analyzing free-text notes, such as natural language processing, could significantly improve detection of these patients to assist hospital epidemiology and infection control efforts.

Disclosures

All Authors: No reported disclosures.

"Should I stay or should I go?" Nurses' perspectives about working during the Covid-19 pandemic's first wave in the United States: A summative content analysis combined with topic modeling

BACKGROUND

The COVID-19 pandemic had its first peak in the United States between April and July of 2020, with incidence and prevalence rates of the virus the greatest in the northeastern coast of the country. At the time of study implementation, there were few studies capturing the perspectives of nurses working the frontlines of the pandemic in any setting as research output in the United States focused largely on treating the disease.

OBJECTIVE

The purpose of this study was to capture the perspectives of nurses in the United States working the frontlines of the COVID-19 pandemic's first wave. We were specifically interested in examining the impact of the pandemic on nurses' roles, professional relationships, and the organizational cultures of their employers.

DESIGN

We conducted an online qualitative study with a pragmatic design to capture the perspectives of nurses working during the first wave of the United States COVID-19 pandemic. Through social networking recruitment, frontline nurses from across the country were invited to participate. Participants provided long form, text-based responses to four questions designed to capture their experiences. A combination of Latent Dirichlet Allocation--a natural language processing technique--along with traditional summative content analysis techniques were used to analyze the data.

SETTING

The United States during the COVID-19 pandemic's first wave between May and July of 2020.

RESULTS

A total of 318 nurses participated from 29 out of 50 states, with 242 fully completing all questions. Findings suggested that the place of work mattered significantly in terms of the frontline working experience. It influenced role changes, risk assumption, interprofessional teamwork experiences, and ultimately, likelihood to leave their jobs or the profession altogether. Organizational culture and its influence on pandemic response implementation was a critical feature of their experiences.

CONCLUSIONS

Findings suggest that organizational performance during the pandemic may be reflected in nursing workforce retention as the risk for workforce attrition appears high. It was also clear from the reports that nurses appear to have assumed higher occupational risks during the pandemic when compared to other providers. The 2020 data from this study also offered a number of signals about potential threats to the stability and sustainability of the US nursing workforce that are now manifesting. The findings underscore the importance of conducting health workforce research during a crisis in order to discern the signals of future problems or for long-term crisis response.

TWEETABLE ABSTRACT

Healthcare leaders made the difference for nurses during the pandemic. How many nurses leave their employer in the next year will tell you who was good, who wasn't.

Parallel processing in speech perception with local and global representations of linguistic context

Speech processing is highly incremental. It is widely accepted that human listeners continuously use the linguistic context to anticipate upcoming concepts, words, and phonemes. However, previous evidence supports two seemingly contradictory models of how a predictive context is integrated with the bottom-up sensory input: Classic psycholinguistic paradigms suggest a two-stage process, in which acoustic input initially leads to local, context-independent representations, which are then quickly integrated with contextual constraints. This contrasts with the view that the brain constructs a single coherent, unified interpretation of the input, which fully integrates available information across representational hierarchies, and thus uses contextual constraints to modulate even the earliest sensory representations. To distinguish these hypotheses, we tested magnetoencephalography responses to continuous narrative speech for signatures of local and unified predictive models. Results provide evidence that listeners employ both types of models in parallel. Two local context models uniquely predict some part of early neural responses, one based on sublexical phoneme sequences, and one based on the phonemes in the current word alone; at the same time, even early responses to phonemes also reflect a unified model that incorporates sentence-level constraints to predict upcoming phonemes. Neural source localization places the anatomical origins of the different predictive models in nonidentical parts of the superior temporal lobes bilaterally, with the right hemisphere showing a relative preference for more local models. These results suggest that speech processing recruits both local and unified predictive models in parallel, reconciling previous disparate findings. Parallel models might make the perceptual system more robust, facilitate processing of unexpected inputs, and serve a function in language acquisition.

Naturally occurring language as a source of evidence in suicide prevention

We discuss computational language analysis as it pertains to suicide prevention research, with an emphasis on providing non-technologists with an understanding of key issues and, equally important, considering its relation to the broader enterprise of suicide prevention. Our emphasis here is on naturally occurring language in social media, motivated by its non-intrusive ability to yield high-value information that in the past has been largely unavailable to clinicians.

Blinded Clinical Ratings of Social Media Data are Correlated with In-Person Clinical Ratings in Participants Diagnosed with Either Depression, Schizophrenia, or Healthy Controls

This study investigates clinically valid signals about psychiatric symptoms in social media data, by rating severity of psychiatric symptoms in donated, de-identified Facebook posts and comparing to in-person clinical assessments. Participants with schizophrenia (N=8), depression (N=7), or who were healthy controls (N=8) also consented to the collection of their Facebook activity from three months before the in-person assessments to six weeks after this evaluation. Depressive symptoms were assessed in- person using the Montgomery-Åsberg Depression Rating Scale (MADRS), psychotic symptoms were assessed using the Brief Psychiatric Rating Scale (BPRS), and global functioning was assessed using the Community Assessment of Psychotic Experiences (CAPE-42). Independent raters (psychiatrists, non-psychiatrist mental health clinicians, and two staff members) rated depression, psychosis, and global functioning symptoms from the social media activity of deidentified participants. The correlations between in-person clinical ratings and blinded ratings based on social media data were evaluated. Significant correlations (and trends for significance in the mixed model controlling for multiple raters) were found for psychotic symptoms, global symptom ratings and depressive symptoms. Results like these, indicating the presence of clinically valid signal in social media, are an important step toward developing computational tools that could assist clinicians by providing additional data outside the context of clinical encounters.

Using targeted paraphrasing and monolingual crowdsourcing to improve translation

Targeted paraphrasing is a new approach to the problem of obtaining cost-effective, reasonable quality translation, which makes use of simple and inexpensive human computations by monolingual speakers in combination with machine translation. The key insight behind the process is that it is possible to spot likely translation errors with only monolingual knowledge of the target language, and it is possible to generate alternative ways to say the same thing (i.e., paraphrases) with only monolingual knowledge of the source language. Formal evaluation demonstrates that this approach can yield substantial improvements in translation quality, and the idea has been integrated into a broader framework for monolingual collaborative translation that produces fully accurate, fully fluent translations for a majority of sentences in a real-world translation task, with no involvement of human bilingual speakers.

Semantic Similarity in a Taxonomy: An Information-Based Measure and its Application to Problems of Ambiguity in Natural Language

This article presents a measure of semantic similarity in an IS-A taxonomy based on the notion of shared information content. Experimental evaluation against a benchmark set of human similarity judgments demonstrates that the measure performs better than the traditional edge-counting approach. The article presents algorithms that take advantage of taxonomic similarity in resolving syntactic and semantic ambiguity, along with experimental results demonstrating their effectiveness.

Selectional constraints: an information-theoretic model and its computational realization

A new, information-theoretic model of selectional constraints is proposed. The strategy adopted here is a minimalist one: how far can one get making as few assumptions as possible? In keeping with that strategy, the proposed model consists of only two components: first, a fairly generic taxonomic representation of concepts, and, second, a probabilistic formalization of selectional constraints defined in terms of that taxonomy, computed on the basis of simple, observable frequencies of co-occurrence between predicates and their arguments. Unlike traditional selection restrictions, the information-theoretic approach avoids empirical problems associated with definitional theories of word meaning, accommodates the observation that semantic anomaly often appears to be a matter of degree, and provides an account of how selectional constraints can be learned. A computational implementation of the model "learns" selectional constraints from collections of naturally occurring text; the predictions of the implemented model are evaluated against judgments elicited from adult subjects, and used to explore the way that arguments are syntactically realized for a class of English verbs. The paper concludes with a discussion of the role of selectional constraints in the acquisition of verb meaning.