Validating emergency department vital signs using a data quality engine for data warehouse

Explanatory argumentation in natural language for correct and incorrect medical diagnoses

BACKGROUND

A huge amount of research is carried out nowadays in Artificial Intelligence to propose automated ways to analyse medical data with the aim to support doctors in delivering medical diagnoses. However, a main issue of these approaches is the lack of transparency and interpretability of the achieved results, making it hard to employ such methods for educational purposes. It is therefore necessary to develop new frameworks to enhance explainability in these solutions.

RESULTS

In this paper, we present a novel full pipeline to generate automatically natural language explanations for medical diagnoses. The proposed solution starts from a clinical case description associated with a list of correct and incorrect diagnoses and, through the extraction of the relevant symptoms and findings, enriches the information contained in the description with verified medical knowledge from an ontology. Finally, the system returns a pattern-based explanation in natural language which elucidates why the correct (incorrect) diagnosis is the correct (incorrect) one. The main contribution of the paper is twofold: first, we propose two novel linguistic resources for the medical domain (i.e, a dataset of 314 clinical cases annotated with the medical entities from UMLS, and a database of biological boundaries for common findings), and second, a full Information Extraction pipeline to extract symptoms and findings from the clinical cases and match them with the terms in a medical ontology and to the biological boundaries. An extensive evaluation of the proposed approach shows the our method outperforms comparable approaches.

CONCLUSIONS

Our goal is to offer AI-assisted educational support framework to form clinical residents to formulate sound and exhaustive explanations for their diagnoses to patients.

Applying Natural Language Processing to Textual Data From Clinical Data Warehouses: Systematic Review

BACKGROUND

In recent years, health data collected during the clinical care process have been often repurposed for secondary use through clinical data warehouses (CDWs), which interconnect disparate data from different sources. A large amount of information of high clinical value is stored in unstructured text format. Natural language processing (NLP), which implements algorithms that can operate on massive unstructured textual data, has the potential to structure the data and make clinical information more accessible.

OBJECTIVE

The aim of this review was to provide an overview of studies applying NLP to textual data from CDWs. It focuses on identifying the (1) NLP tasks applied to data from CDWs and (2) NLP methods used to tackle these tasks.

METHODS

This review was performed according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. We searched for relevant articles in 3 bibliographic databases: PubMed, Google Scholar, and ACL Anthology. We reviewed the titles and abstracts and included articles according to the following inclusion criteria: (1) focus on NLP applied to textual data from CDWs, (2) articles published between 1995 and 2021, and (3) written in English.

RESULTS

We identified 1353 articles, of which 194 (14.34%) met the inclusion criteria. Among all identified NLP tasks in the included papers, information extraction from clinical text (112/194, 57.7%) and the identification of patients (51/194, 26.3%) were the most frequent tasks. To address the various tasks, symbolic methods were the most common NLP methods (124/232, 53.4%), showing that some tasks can be partially achieved with classical NLP techniques, such as regular expressions or pattern matching that exploit specialized lexica, such as drug lists and terminologies. Machine learning (70/232, 30.2%) and deep learning (38/232, 16.4%) have been increasingly used in recent years, including the most recent approaches based on transformers. NLP methods were mostly applied to English language data (153/194, 78.9%).

CONCLUSIONS

CDWs are central to the secondary use of clinical texts for research purposes. Although the use of NLP on data from CDWs is growing, there remain challenges in this field, especially with regard to languages other than English. Clinical NLP is an effective strategy for accessing, extracting, and transforming data from CDWs. Information retrieved with NLP can assist in clinical research and have an impact on clinical practice.

Identification and Prediction of Clinical Phenotypes in Hospitalized Patients With COVID-19: Machine Learning From Medical Records

BACKGROUND

There is significant heterogeneity in disease progression among hospitalized patients with COVID-19. The pathogenesis of SARS-CoV-2 infection is attributed to a complex interplay between virus and host immune response that in some patients unpredictably and rapidly leads to "hyperinflammation" associated with increased risk of mortality. The early identification of patients at risk of progression to hyperinflammation may help inform timely therapeutic decisions and lead to improved outcomes.

OBJECTIVE

The primary objective of this study was to use machine learning to reproducibly identify specific risk-stratifying clinical phenotypes across hospitalized patients with COVID-19 and compare treatment response characteristics and outcomes. A secondary objective was to derive a predictive phenotype classification model using routinely available early encounter data that may be useful in informing optimal COVID-19 bedside clinical management.

METHODS

This was a retrospective analysis of electronic health record data of adult patients (N=4379) who were admitted to a Johns Hopkins Health System hospital for COVID-19 treatment from 2020 to 2021. Phenotypes were identified by clustering 38 routine clinical observations recorded during inpatient care. To examine the reproducibility and validity of the derived phenotypes, patient data were randomly divided into 2 cohorts, and clustering analysis was performed independently for each cohort. A predictive phenotype classifier using the gradient-boosting machine method was derived using routine clinical observations recorded during the first 6 hours following admission.

RESULTS

A total of 2 phenotypes (designated as phenotype 1 and phenotype 2) were identified in patients admitted for COVID-19 in both the training and validation cohorts with similar distributions of features, correlations with biomarkers, treatments, comorbidities, and outcomes. In both the training and validation cohorts, phenotype-2 patients were older; had elevated markers of inflammation; and were at an increased risk of requiring intensive care unit-level care, developing sepsis, and mortality compared with phenotype-1 patients. The gradient-boosting machine phenotype prediction model yielded an area under the curve of 0.89 and a positive predictive value of 0.83.

CONCLUSIONS

Using machine learning clustering, we identified and internally validated 2 clinical COVID-19 phenotypes with distinct treatment or response characteristics consistent with similar 2-phenotype models derived from other hospitalized populations with COVID-19, supporting the reliability and generalizability of these findings. COVID-19 phenotypes can be accurately identified using machine learning models based on readily available early encounter clinical data. A phenotype prediction model based on early encounter data may be clinically useful for timely bedside risk stratification and treatment personalization.

The Use and Structure of Emergency Nurses' Triage Narrative Data: Scoping Review

BACKGROUND

Emergency departments use triage to ensure that patients with the highest level of acuity receive care quickly and safely. Triage is typically a nursing process that is documented as structured and unstructured (free text) data. Free-text triage narratives have been studied for specific conditions but never reviewed in a comprehensive manner.

OBJECTIVE

The objective of this paper was to identify and map the academic literature that examines triage narratives. The paper described the types of research conducted, identified gaps in the research, and determined where additional review may be warranted.

METHODS

We conducted a scoping review of unstructured triage narratives. We mapped the literature, described the use of triage narrative data, examined the information available on the form and structure of narratives, highlighted similarities among publications, and identified opportunities for future research.

RESULTS

We screened 18,074 studies published between 1990 and 2022 in CINAHL, MEDLINE, Embase, Cochrane, and ProQuest Central. We identified 0.53% (96/18,074) of studies that directly examined the use of triage nurses' narratives. More than 12 million visits were made to 2438 emergency departments included in the review. In total, 82% (79/96) of these studies were conducted in the United States (43/96, 45%), Australia (31/96, 32%), or Canada (5/96, 5%). Triage narratives were used for research and case identification, as input variables for predictive modeling, and for quality improvement. Overall, 31% (30/96) of the studies offered a description of the triage narrative, including a list of the keywords used (27/96, 28%) or more fulsome descriptions (such as word counts, character counts, abbreviation, etc; 7/96, 7%). We found limited use of reporting guidelines (8/96, 8%).

CONCLUSIONS

The breadth of the identified studies suggests that there is widespread routine collection and research use of triage narrative data. Despite the use of triage narratives as a source of data in studies, the narratives and nurses who generate them are poorly described in the literature, and data reporting is inconsistent. Additional research is needed to describe the structure of triage narratives, determine the best use of triage narratives, and improve the consistent use of triage-specific data reporting guidelines.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

RR2-10.1136/bmjopen-2021-055132.

Effect of Deintensifying Diabetes Medications on Negative Events in Older Veteran Nursing Home Residents

OBJECTIVE

Guidelines advocate against tight glycemic control in older nursing home (NH) residents with advanced dementia (AD) or limited life expectancy (LLE). We evaluated the effect of deintensifying diabetes medications with regard to all-cause emergency department (ED) visits, hospitalizations, and death in NH residents with LLE/AD and tight glycemic control.

RESEARCH DESIGN AND METHODS

We conducted a national retrospective cohort study of 2,082 newly admitted nonhospice veteran NH residents with LLE/AD potentially overtreated for diabetes (HbA1c ≤7.5% and one or more diabetes medications) in fiscal years 2009-2015. Diabetes treatment deintensification (dose decrease or discontinuation of a noninsulin agent or stopping insulin sustained ≥7 days) was identified within 30 days after HbA1c measurement. To adjust for confounding, we used entropy weights to balance covariates between NH residents who deintensified versus continued medications. We used the Aalen-Johansen estimator to calculate the 60-day cumulative incidence and risk ratios (RRs) for ED or hospital visits and deaths.

RESULTS

Diabetes medications were deintensified for 27% of residents. In the subsequent 60 days, 28.5% of all residents were transferred to the ED or acute hospital setting for any cause and 3.9% died. After entropy weighting, deintensifying was not associated with 60-day all-cause ED visits or hospitalizations (RR 0.99 [95% CI 0.84, 1.18]) or 60-day mortality (1.52 [0.89, 2.81]).

CONCLUSIONS

Among NH residents with LLE/AD who may be inappropriately overtreated with tight glycemic control, deintensification of diabetes medications was not associated with increased risk of 60-day all-cause ED visits, hospitalization, or death.

An exploratory data quality analysis of time series physiologic signals using a large-scale intensive care unit database

Physiological data, such as heart rate and blood pressure, are critical to clinical decision-making in the intensive care unit (ICU). Vital signs data, which are available from electronic health records, can be used to diagnose and predict important clinical outcomes; While there have been some reports on the data quality of nurse-verified vital sign data, little has been reported on the data quality of higher frequency time-series vital signs acquired in ICUs, that would enable such predictive modeling. In this study, we assessed the data quality issues, defined as the completeness, accuracy, and timeliness, of minute-by-minute time series vital signs data within the MIMIC-III data set, captured from 16009 patient-ICU stays and corresponding to 9410 unique adult patients. We measured data quality of four time-series vital signs data streams in the MIMIC-III data set: heart rate (HR), respiratory rate (RR), blood oxygen saturation (SpO2), and arterial blood pressure (ABP). Approximately, 30% of patient-ICU stays did not have at least 1 min of data during the time-frame of the ICU stay for HR, RR, and SpO2. The percentage of patient-ICU stays that did not have at least 1 min of ABP data was ∼56%. We observed ∼80% coverage of the total duration of the ICU stay for HR, RR, and SpO2. Finally, only 12.5%%, 9.9%, 7.5%, and 4.4% of ICU lengths of stay had ≥ 99% data available for HR, RR, SpO2, and ABP, respectively, that would meet the three data quality requirements we looked into in this study. Our findings on data completeness, accuracy, and timeliness have important implications for data scientists and informatics researchers who use time series vital signs data to develop predictive models of ICU outcomes.

Impact of corticosteroids in hospitalised COVID-19 patients

BACKGROUND

Corticosteroids are a potential therapeutic agent for patients with COVID-19 pneumonia. The RECOVERY (Randomised Trials in COVID-19 Therapy) trial provided data on the mortality benefits of corticosteroids. The study aimed to determine the association between corticosteroid use on mortality and infection rates and to define subgroups who may benefit from corticosteroids in a real-world setting.

METHODS

Clinical data were extracted that included demographic, laboratory data and details of the therapy, including the administration of corticosteroids, azithromycin, hydroxychloroquine, tocilizumab and anticoagulation. The primary outcome was in-hospital mortality. Secondary outcomes included intensive care unit (ICU) admission and invasive mechanical ventilation. Outcomes were compared in patients who did and did not receive corticosteroids using the multivariate Cox regression model.

RESULTS

4313 patients were hospitalised with COVID-19 during the study period, of whom 1270 died (29.4%). When administered within the first 7 days after admission, corticosteroids were associated with reduced mortality (OR 0.73, 95% CI 0.55 to 0.97, p=0.03) and decreased transfers to the ICU (OR 0.72, 95% CI 0.47 to 1.11, p=0.02). This mortality benefit was particularly impressive in younger patients (<65 years of age), females and those with elevated inflammatory markers, defined as C reactive protein ≥150 mg/L (p≤0.05), interleukin-6 ≥20 pg/mL (p≤0.05) or D-dimer ≥2.0 µg/L (p≤0.05). Therapy was safe with similar rates of bacteraemia and fungaemia in corticosteroid-treated and non-corticosteroid-treated patients.

CONCLUSION

In patients hospitalised with COVID-19 pneumonia, corticosteroid use within the first 7 days of admission decreased mortality and ICU admissions with no associated increase in bacteraemia or fungaemia.

Antihypertensive Deprescribing in Older Adult Veterans at End of Life Admitted to Veteran Affairs Nursing Homes

OBJECTIVES

Geriatric palliative care approaches support deprescribing of antihypertensives in older nursing home (NH) residents with limited life expectancy and/or advanced dementia (LLE/AD) who are intensely treated for hypertension (HTN), but information on real-world deprescribing patterns in this population is limited. We examined the incidence and factors associated with antihypertensive deprescribing.

DESIGN

National, retrospective cohort study.

SETTING AND PARTICIPANTS

Older Veterans with LLE/AD and HTN admitted to VA NHs in fiscal years 2009-2015 with potential overtreatment of HTN at admission, defined as receiving at least 1 antihypertensive class of medications and mean daily systolic blood pressure (SBP) <120 mm Hg.

MEASURES

Deprescribing was defined as subsequent dose reduction or discontinuation of an antihypertensive for ≥7 days. Competing risk models assessed cumulative incidence and factors associated with deprescribing.

RESULTS

Within our sample (n = 10,574), cumulative incidence of deprescribing at 30 days was 41%. Veterans with the greatest level of overtreatment (ie, multiple antihypertensives and SBP <100 mm Hg) had an increased likelihood (hazard ratio 1.75, 95% confidence interval 1.59, 1.93) of deprescribing vs those with the lowest level of overtreatment (ie, one antihypertensive and SBP ≥100 to <120 mm Hg). Several markers of poor prognosis (ie, recent weight loss, poor appetite, dehydration, dependence for activities of daily living, pain) and later admission year were associated with increased likelihood of deprescribing, whereas cardiovascular risk factors (ie, diabetes, congestive heart failure, obesity), shortness of breath, and admission source from another NH or home/assisted living setting (vs acute hospital) were associated with decreased likelihood.

CONCLUSIONS AND IMPLICATIONS

Real-world deprescribing patterns of antihypertensives among NH residents with HTN and LLE/AD appear to reflect variation in recommendations for HTN treatment intensity and individualization of patient care in a population with potential overtreatment. Factors facilitating deprescribing included treatment intensity and markers of poor prognosis. Comparative effectiveness and safety studies are needed to guide clinical decisions around deprescribing and HTN management.

The impact of windows and daylight on acute-care nurses' physiological, psychological, and behavioral health

OBJECTIVE

To investigate the physiological and psychological effects of windows and daylight on registered nurses.

BACKGROUND

To date, evidence has indicated that appropriate environmental lighting with characteristics similar to natural light can improve mood, alertness, and performance. The restorative effects of windows also have been documented. Hospital workspaces generally lack windows and daylight, and the impact of the lack of windows and daylight on healthcare employees' well being has not been thoroughly investigated.

METHODS

Data were collected using multiple methods with a quasi-experimental approach (i.e., biological measurements, behavioral mapping, and analysis of archival data) in an acute-care nursing unit with two wards that have similar environmental and organizational conditions, and similar patient populations and acuity, but different availability of windows in the nursing stations.

RESULTS

Findings indicated that blood pressure (p < 0.0001) decreased and body temperature increased (p = 0.03). Blood oxygen saturation increased (p = 0.02), but the difference was clinically insignificant. Communication (p < 0.0001) and laughter (p = 0.03) both increased, and the subsidiary behavior indicators of sleepiness and deteriorated mood (p = 0.02) decreased. Heart rate (p = 0.07), caffeine intake (p = 0.3), self-reported sleepiness (p = 0.09), and the frequency of medication errors (p = 0.14) also decreased, but insignificantly.

CONCLUSIONS

The findings support evidence from laboratory and field settings of the benefits of windows and daylight. A possible micro-restorative effect of windows and daylight may result in lowered blood pressure and increased oxygen saturation and a positive effect on circadian rhythms (as suggested by body temperature) and morning sleepiness.

KEYWORDS

Critical care/intensive care, lighting, nursing, quality care, work environment.

A Data Quality Assessment Guideline for Electronic Health Record Data Reuse

Introduction:

We describe the formulation, development, and initial expert review of 3x3 Data Quality Assessment (DQA), a dynamic, evidence-based guideline to enable electronic health record (EHR) data quality assessment and reporting for clinical research.

Methods:

3x3 DQA was developed through the triangulation results from three studies: a review of the literature on EHR data quality assessment, a quantitative study of EHR data completeness, and a set of interviews with clinical researchers. Following initial development, the guideline was reviewed by a panel of EHR data quality experts.

Results:

The guideline embraces the task-dependent nature of data quality and data quality assessment. The core framework includes three constructs of data quality: complete, correct, and current data. These constructs are operationalized according to the three primary dimensions of EHR data: patients, variables, and time. Each of the nine operationalized constructs maps to a methodological recommendation for EHR data quality assessment. The initial expert response to the framework was positive, but improvements are required.

Discussion:

The initial version of 3x3 DQA promises to enable explicit guideline-based best practices for EHR data quality assessment and reporting. Future work will focus on increasing clarity on how and when 3x3 DQA should be used during the research process, improving the feasibility and ease-of-use of recommendation execution, and clarifying the process for users to determine which operationalized constructs and recommendations are relevant for a given dataset and study.

Exploring Vital Sign Data Quality in Electronic Health Records with Focus on Emergency Care Warning Scores

BACKGROUND

Computerized clinical decision support and automation of warnings have been advocated to assist clinicians in detecting patients at risk of physiological instability. To provide reliable support such systems are dependent on high-quality vital sign data. Data quality depends on how, when and why the data is captured and/or documented.

OBJECTIVES

This study aims to describe the effects on data quality of vital signs by three different types of documentation practices in five Swedish emergency hospitals, and to assess data fitness for calculating warning and triage scores. The study also provides reference data on triage vital signs in Swedish emergency care.

METHODS

We extracted a dataset including vital signs, demographic and administrative data from emergency care visits (n=335027) at five Swedish emergency hospitals during 2013 using either completely paper-based, completely electronic or mixed documentation practices. Descriptive statistics were used to assess fitness for use in emergency care decision support systems aiming to calculate warning and triage scores, and data quality was described in three categories: currency, completeness and correctness. To estimate correctness, two further categories - plausibility and concordance - were used.

RESULTS

The study showed an acceptable correctness of the registered vital signs irrespectively of the type of documentation practice. Completeness was high in sites where registrations were routinely entered into the Electronic Health Record (EHR). The currency was only acceptable in sites with a completely electronic documentation practice.

CONCLUSION

Although vital signs that were recorded in completely electronic documentation practices showed plausible results regarding correctness, completeness and currency, the study concludes that vital signs documented in Swedish emergency care EHRs cannot generally be considered fit for use for calculation of triage and warning scores. Low completeness and currency were found if the documentation was not completely electronic.

Empirical advances with text mining of electronic health records

Background

Korian is a private group specializing in medical accommodations for elderly and dependent people. A professional data warehouse (DWH) established in 2010 hosts all of the residents’ data. Inside this information system (IS), clinical narratives (CNs) were used only by medical staff as a residents’ care linking tool.

The objective of this study was to show that, through qualitative and quantitative textual analysis of a relatively small physiotherapy and well-defined CN sample, it was possible to build a physiotherapy corpus and, through this process, generate a new body of knowledge by adding relevant information to describe the residents’ care and lives.

Methods

Meaningful words were extracted through Standard Query Language (SQL) with the LIKE function and wildcards to perform pattern matching, followed by text mining and a word cloud using R® packages. Another step involved principal components and multiple correspondence analyses, plus clustering on the same residents’ sample as well as on other health data using a health model measuring the residents’ care level needs.

Results

By combining these techniques, physiotherapy treatments could be characterized by a list of constructed keywords, and the residents’ health characteristics were built. Feeding defects or health outlier groups could be detected, physiotherapy residents’ data and their health data were matched, and differences in health situations showed qualitative and quantitative differences in physiotherapy narratives.

Conclusions

This textual experiment using a textual process in two stages showed that text mining and data mining techniques provide convenient tools to improve residents’ health and quality of care by adding new, simple, useable data to the electronic health record (EHR). When used with a normalized physiotherapy problem list, text mining through information extraction (IE), named entity recognition (NER) and data mining (DM) can provide a real advantage to describe health care, adding new medical material and helping to integrate the EHR system into the health staff work environment.

Electronic supplementary material

The online version of this article (doi:10.1186/s12911-017-0519-0) contains supplementary material, which is available to authorized users.

Text mining approach to predict hospital admissions using early medical records from the emergency department

OBJECTIVE

Emergency department (ED) overcrowding is a serious issue for hospitals. Early information on short-term inward bed demand from patients receiving care at the ED may reduce the overcrowding problem, and optimize the use of hospital resources. In this study, we use text mining methods to process data from early ED patient records using the SOAP framework, and predict future hospitalizations and discharges.

DESIGN

We try different approaches for pre-processing of text records and to predict hospitalization. Sets-of-words are obtained via binary representation, term frequency, and term frequency-inverse document frequency. Unigrams, bigrams and trigrams are tested for feature formation. Feature selection is based on χ² and F-score metrics. In the prediction module, eight text mining methods are tested: Decision Tree, Random Forest, Extremely Randomized Tree, AdaBoost, Logistic Regression, Multinomial Naïve Bayes, Support Vector Machine (Kernel linear) and Nu-Support Vector Machine (Kernel linear).

MEASUREMENTS

Prediction performance is evaluated by F1-scores. Precision and Recall values are also informed for all text mining methods tested.

RESULTS

Nu-Support Vector Machine was the text mining method with the best overall performance. Its average F1-score in predicting hospitalization was 77.70%, with a standard deviation (SD) of 0.66%.

CONCLUSIONS

The method could be used to manage daily routines in EDs such as capacity planning and resource allocation. Text mining could provide valuable information and facilitate decision-making by inward bed management teams.

How to improve vital sign data quality for use in clinical decision support systems? A qualitative study in nine Swedish emergency departments

Background

Vital sign data are important for clinical decision making in emergency care. Clinical Decision Support Systems (CDSS) have been advocated to increase patient safety and quality of care. However, the efficiency of CDSS depends on the quality of the underlying vital sign data. Therefore, possible factors affecting vital sign data quality need to be understood.

This study aims to explore the factors affecting vital sign data quality in Swedish emergency departments and to determine in how far clinicians perceive vital sign data to be fit for use in clinical decision support systems. A further aim of the study is to provide recommendations on how to improve vital sign data quality in emergency departments.

Methods

Semi-structured interviews were conducted with sixteen physicians and nurses from nine hospitals and vital sign documentation templates were collected and analysed. Follow-up interviews and process observations were done at three of the hospitals to verify the results. Content analysis with constant comparison of the data was used to analyse and categorize the collected data.

Results

Factors related to care process and information technology were perceived to affect vital sign data quality. Despite electronic health records (EHRs) being available in all hospitals, these were not always used for vital sign documentation. Only four out of nine sites had a completely digitalized vital sign documentation flow and paper-based triage records were perceived to provide a better mobile workflow support than EHRs. Observed documentation practices resulted in low currency, completeness, and interoperability of the vital signs. To improve vital sign data quality, we propose to standardize the care process, improve the digital documentation support, provide workflow support, ensure interoperability and perform quality control.

Conclusions

Vital sign data quality in Swedish emergency departments is currently not fit for use by CDSS. To address both technical and organisational challenges, we propose five steps for vital sign data quality improvement to be implemented in emergency care settings.

Electronic supplementary material

The online version of this article (doi:10.1186/s12911-016-0305-4) contains supplementary material, which is available to authorized users.