%0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e20071 %T Improving Diagnostic Classification of Stillbirths and Neonatal Deaths Using ICD-PM (International Classification of Diseases for Perinatal Mortality) Codes: Validation Study %A Luk,Hiu Mei %A Allanson,Emma %A Ming,Wai-Kit %A Leung,Wing Cheong %+ Department of Obstetrics and Gynaecology, Kwong Wah Hospital, Hong Kong SAR, China (Hong Kong), 852 35175091, leungwc@ha.org.hk %K stillbirths %K perinatal deaths %K neonatal deaths %K ICD-PM %K ICD-10 %D 2020 %7 3.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: Stillbirths and neonatal deaths have long been imperfectly classified and recorded worldwide. In Hong Kong, the current code system is deficient (>90% cases with unknown causes) in providing the diagnoses of perinatal mortality cases. Objective: The objective of this study was to apply the International Classification of Diseases for Perinatal Mortality (ICD-PM) system to existing perinatal death data. Further, the aim was to assess whether there was any change in the classifications of perinatal deaths compared with the existing classification system and identify any areas in which future interventions can be made. Methods: We applied the ICD-PM (with International Statistical Classification of Diseases and Related Health Problems, 10th Revision) code system to existing perinatal death data in Kwong Wah Hospital, Hong Kong, to improve diagnostic classification. The study included stillbirths (after 24 weeks gestation) and neonatal deaths (from birth to 28 days). The retrospective data (5 years) from May 1, 2012, to April 30, 2017, were recoded by the principal investigator (HML) applying the ICD-PM, then validated by an overseas expert (EA) after she reviewed the detailed case summaries. The prospective application of ICD-PM from May 1, 2017, to April 30, 2019, was performed during the monthly multidisciplinary perinatal meetings and then also validated by EA for agreement. Results: We analyzed the data of 34,920 deliveries, and 119 cases were included for analysis (92 stillbirths and 27 neonatal deaths). The overall agreement with EA of our codes using the ICD-PM was 93.2% (111/119); 92% (78/85) for the 5 years of retrospective codes and 97% (33/34) for the 2 years of prospective codes (P=.44). After the application of the ICD-PM, the overall proportion of unknown causes of perinatal mortality dropped from 34.5% (41/119) to 10.1% (12/119) of cases (P<.001). Conclusions: Using the ICD-PM would lead to a better classification of perinatal deaths, reduce the proportion of unknown diagnoses, and clearly link the maternal conditions with these perinatal deaths. %M 32744510 %R 10.2196/20071 %U https://medinform.jmir.org/2020/8/e20071 %U https://doi.org/10.2196/20071 %U http://www.ncbi.nlm.nih.gov/pubmed/32744510 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e21056 %T Impact of a Commercial Artificial Intelligence–Driven Patient Self-Assessment Solution on Waiting Times at General Internal Medicine Outpatient Departments: Retrospective Study %A Harada,Yukinori %A Shimizu,Taro %+ Department of Diagnostic and Generalist Medicine, Dokkyo Medical University, Kitakobayashi 880, Mibu, 321-0293, Japan, 81 282861111, shimizutaro7@gmail.com %K artificial intelligence %K automated medical history taking system %K eHealth %K interrupted time-series analysis %K waiting time %D 2020 %7 31.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: Patient waiting time at outpatient departments is directly related to patient satisfaction and quality of care, particularly in patients visiting the general internal medicine outpatient departments for the first time. Moreover, reducing wait time from arrival in the clinic to the initiation of an examination is key to reducing patients’ anxiety. The use of automated medical history–taking systems in general internal medicine outpatient departments is a promising strategy to reduce waiting times. Recently, Ubie Inc in Japan developed AI Monshin, an artificial intelligence–based, automated medical history–taking system for general internal medicine outpatient departments. Objective: We hypothesized that replacing the use of handwritten self-administered questionnaires with the use of AI Monshin would reduce waiting times in general internal medicine outpatient departments. Therefore, we conducted this study to examine whether the use of AI Monshin reduced patient waiting times. Methods: We retrospectively analyzed the waiting times of patients visiting the general internal medicine outpatient department at a Japanese community hospital without an appointment from April 2017 to April 2020. AI Monshin was implemented in April 2019. We compared the median waiting time before and after implementation by conducting an interrupted time-series analysis of the median waiting time per month. We also conducted supplementary analyses to explain the main results. Results: We analyzed 21,615 visits. The median waiting time after AI Monshin implementation (74.4 minutes, IQR 57.1) was not significantly different from that before AI Monshin implementation (74.3 minutes, IQR 63.7) (P=.12). In the interrupted time-series analysis, the underlying linear time trend (–0.4 minutes per month; P=.06; 95% CI –0.9 to 0.02), level change (40.6 minutes; P=.09; 95% CI –5.8 to 87.0), and slope change (–1.1 minutes per month; P=.16; 95% CI –2.7 to 0.4) were not statistically significant. In a supplemental analysis of data from 9054 of 21,615 visits (41.9%), the median examination time after AI Monshin implementation (6.0 minutes, IQR 5.2) was slightly but significantly longer than that before AI Monshin implementation (5.7 minutes, IQR 5.0) (P=.003). Conclusions: The implementation of an artificial intelligence–based, automated medical history–taking system did not reduce waiting time for patients visiting the general internal medicine outpatient department without an appointment, and there was a slight increase in the examination time after implementation; however, the system may have enhanced the quality of care by supporting the optimization of staff assignments. %M 32865504 %R 10.2196/21056 %U http://medinform.jmir.org/2020/8/e21056/ %U https://doi.org/10.2196/21056 %U http://www.ncbi.nlm.nih.gov/pubmed/32865504 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e20359 %T Utilization Barriers and Medical Outcomes Commensurate With the Use of Telehealth Among Older Adults: Systematic Review %A Kruse,Clemens %A Fohn,Joanna %A Wilson,Nakia %A Nunez Patlan,Evangelina %A Zipp,Stephanie %A Mileski,Michael %+ School of Health Administration, Texas State University, 601 University Dr, Encino Hall 250, San Marcos, TX, 78666, United States, 1 210 355 4742, scottkruse@txstate.edu %K telehealth %K telemedicine %K older adults %K barriers %K health outcomes %D 2020 %7 12.8.2020 %9 Review %J JMIR Med Inform %G English %X Background: Rising telehealth capabilities and improving access to older adults can aid in improving health outcomes and quality of life indicators. Telehealth is not being used ubiquitously at present. Objective: This review aimed to identify the barriers that prevent ubiquitous use of telehealth and the ways in which telehealth improves health outcomes and quality of life indicators for older adults. Methods: This systematic review was conducted and reported in accordance with the Kruse protocol and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Reviewers queried the following four research databases: Cumulative Index of Nursing and Allied Health Literature (CINAHL), PubMed (MEDLINE), Web of Science, and Embase (Science Direct). Reviewers analyzed 57 articles, performed a narrative analysis to identify themes, and identified barriers and reports of health outcomes and quality of life indicators found in the literature. Results: Reviewers analyzed 57 studies across the following five interventions of telehealth: eHealth, mobile health (mHealth), telemonitoring, telecare (phone), and telehealth video calls, with a Cohen κ of 0.75. Reviewers identified 14 themes for barriers. The most common of which were technical literacy (25/144 occurrences, 17%), lack of desire (19/144 occurrences, 13%), and cost (11/144 occurrences, 8%). Reviewers identified 13 medical outcomes associated with telehealth interventions. The most common of which were decrease in psychological stress (21/118 occurrences, 18%), increase in autonomy (18/118 occurrences, 15%), and increase in cognitive ability (11/118 occurrences, 9%). Some articles did not report medical outcomes (18/57, 32%) and some did not report barriers (19/57, 33%). Conclusions: The literature suggests that the elimination of barriers could increase the prevalence of telehealth use by older adults. By increasing use of telehealth, proximity to care is no longer an issue for access, and thereby care can reach populations with chronic conditions and mobility restrictions. Future research should be conducted on methods for personalizing telehealth in older adults before implementation. Trial Registration: PROSPERO CRD42020182162; https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42020182162. International Registered Report Identifier (IRRID): RR2-10.2196/15490 %M 32784177 %R 10.2196/20359 %U http://medinform.jmir.org/2020/8/e20359/ %U https://doi.org/10.2196/20359 %U http://www.ncbi.nlm.nih.gov/pubmed/32784177 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e17283 %T Effects of Computerized Decision Support Systems on Practitioner Performance and Patient Outcomes: Systematic Review %A Kruse,Clemens Scott %A Ehrbar,Nolan %+ School of Health Administration, Texas State University, 601 University Dr, San Marcos, TX, , United States, 1 512 245 4462, scottkruse@txstate.edu %K CDSS %K performance %K outcomes %D 2020 %7 11.8.2020 %9 Review %J JMIR Med Inform %G English %X Background: Computerized decision support systems (CDSSs) are software programs that support the decision making of practitioners and other staff. Other reviews have analyzed the relationship between CDSSs, practitioner performance, and patient outcomes. These reviews reported positive practitioner performance in over half the articles analyzed, but very little information was found for patient outcomes. Objective: The purpose of this review was to analyze the relationship between CDSSs, practitioner performance, and patient medical outcomes. PubMed, CINAHL, Embase, Web of Science, and Cochrane databases were queried. Methods: Articles were chosen based on year published (last 10 years), high quality, peer-reviewed sources, and discussion of the relationship between the use of CDSS as an intervention and links to practitioner performance or patient outcomes. Reviewers used an Excel spreadsheet (Microsoft Corporation) to collect information on the relationship between CDSSs and practitioner performance or patient outcomes. Reviewers also collected observations of participants, intervention, comparison with control group, outcomes, and study design (PICOS) along with those showing implicit bias. Articles were analyzed by multiple reviewers following the Kruse protocol for systematic reviews. Data were organized into multiple tables for analysis and reporting. Results: Themes were identified for both practitioner performance (n=38) and medical outcomes (n=36). A total of 66% (25/38) of articles had occurrences of positive practitioner performance, 13% (5/38) found no difference in practitioner performance, and 21% (8/38) did not report or discuss practitioner performance. Zero articles reported negative practitioner performance. A total of 61% (22/36) of articles had occurrences of positive patient medical outcomes, 8% (3/36) found no statistically significant difference in medical outcomes between intervention and control groups, and 31% (11/36) did not report or discuss medical outcomes. Zero articles found negative patient medical outcomes attributed to using CDSSs. Conclusions: Results of this review are commensurate with previous reviews with similar objectives, but unlike these reviews we found a high level of reporting of positive effects on patient medical outcomes. %M 32780714 %R 10.2196/17283 %U http://medinform.jmir.org/2020/8/e17283/ %U https://doi.org/10.2196/17283 %U http://www.ncbi.nlm.nih.gov/pubmed/32780714 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e20974 %T Diagnostic Model for In-Hospital Bleeding in Patients with Acute ST-Segment Elevation Myocardial Infarction: Algorithm Development and Validation %A Li,Yong %+ Emergency and Critical Care Center, Beijing Anzhen Hospital, Capital Medical University, No. 405, Building No. 5, Madian Nancun, Xicheng District, Beijing, 100088, China, 86 13910227262, liyongdoctor@sina.com %K coronary disease %K ST-segment elevation myocardial infarction %K hemorrhage %K nomogram %D 2020 %7 14.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: Bleeding complications in patients with acute ST-segment elevation myocardial infarction (STEMI) have been associated with increased risk of subsequent adverse consequences. Objective: The objective of our study was to develop and externally validate a diagnostic model of in-hospital bleeding. Methods: We performed multivariate logistic regression of a cohort for hospitalized patients with acute STEMI in the emergency department of a university hospital. Participants: The model development data set was obtained from 4262 hospitalized patients with acute STEMI from January 2002 to December 2013. A set of 6015 hospitalized patients with acute STEMI from January 2014 to August 2019 were used for external validation. We used logistic regression analysis to analyze the risk factors of in-hospital bleeding in the development data set. We developed a diagnostic model of in-hospital bleeding and constructed a nomogram. We assessed the predictive performance of the diagnostic model in the validation data sets by examining measures of discrimination, calibration, and decision curve analysis (DCA). Results: In-hospital bleeding occurred in 112 of 4262 participants (2.6%) in the development data set. The strongest predictors of in-hospital bleeding were advanced age and high Killip classification. Logistic regression analysis showed differences between the groups with and without in-hospital bleeding in age (odds ratio [OR] 1.047, 95% CI 1.029-1.066; P<.001), Killip III (OR 3.265, 95% CI 2.008-5.31; P<.001), and Killip IV (OR 5.133, 95% CI 3.196-8.242; P<.001). We developed a diagnostic model of in-hospital bleeding. The area under the receiver operating characteristic curve (AUC) was 0.777 (SD 0.021, 95% CI 0.73576-0.81823). We constructed a nomogram based on age and Killip classification. In-hospital bleeding occurred in 117 of 6015 participants (1.9%) in the validation data set. The AUC was 0.7234 (SD 0.0252, 95% CI 0.67392-0.77289). Conclusions: We developed and externally validated a diagnostic model of in-hospital bleeding in patients with acute STEMI. The discrimination, calibration, and DCA of the model were found to be satisfactory. Trial Registration: ChiCTR.org ChiCTR1900027578; http://www.chictr.org.cn/showprojen.aspx?proj=45926 %M 32795995 %R 10.2196/20974 %U http://medinform.jmir.org/2020/8/e20974/ %U https://doi.org/10.2196/20974 %U http://www.ncbi.nlm.nih.gov/pubmed/32795995 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e18542 %T Model-Based Algorithms for Detecting Peripheral Artery Disease Using Administrative Data From an Electronic Health Record Data System: Algorithm Development Study %A Weissler,Elizabeth Hope %A Lippmann,Steven J %A Smerek,Michelle M %A Ward,Rachael A %A Kansal,Aman %A Brock,Adam %A Sullivan,Robert C %A Long,Chandler %A Patel,Manesh R %A Greiner,Melissa A %A Hardy,N Chantelle %A Curtis,Lesley H %A Jones,W Schuyler %+ Department of Medicine, Duke University School of Medicine, DUMC Box 3330, Durham, NC, 27710, United States, 1 919 668 8917, schuyler.jones@duke.edu %K peripheral artery disease %K patient selection %K electronic health records %K cardiology %K health data %D 2020 %7 19.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: Peripheral artery disease (PAD) affects 8 to 10 million Americans, who face significantly elevated risks of both mortality and major limb events such as amputation. Unfortunately, PAD is relatively underdiagnosed, undertreated, and underresearched, leading to wide variations in treatment patterns and outcomes. Efforts to improve PAD care and outcomes have been hampered by persistent difficulties identifying patients with PAD for clinical and investigatory purposes. Objective: The aim of this study is to develop and validate a model-based algorithm to detect patients with peripheral artery disease (PAD) using data from an electronic health record (EHR) system. Methods: An initial query of the EHR in a large health system identified all patients with PAD-related diagnosis codes for any encounter during the study period. Clinical adjudication of PAD diagnosis was performed by chart review on a random subgroup. A binary logistic regression to predict PAD was built and validated using a least absolute shrinkage and selection operator (LASSO) approach in the adjudicated patients. The algorithm was then applied to the nonsampled records to further evaluate its performance. Results: The initial EHR data query using 406 diagnostic codes yielded 15,406 patients. Overall, 2500 patients were randomly selected for ground truth PAD status adjudication. In the end, 108 code flags remained after removing rarely- and never-used codes. We entered these code flags plus administrative encounter, imaging, procedure, and specialist flags into a LASSO model. The area under the curve for this model was 0.862. Conclusions: The algorithm we constructed has two main advantages over other approaches to the identification of patients with PAD. First, it was derived from a broad population of patients with many different PAD manifestations and treatment pathways across a large health system. Second, our model does not rely on clinical notes and can be applied in situations in which only administrative billing data (eg, large administrative data sets) are available. A combination of diagnosis codes and administrative flags can accurately identify patients with PAD in large cohorts. %M 32663152 %R 10.2196/18542 %U http://medinform.jmir.org/2020/8/e18542/ %U https://doi.org/10.2196/18542 %U http://www.ncbi.nlm.nih.gov/pubmed/32663152 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e18389 %T Analysis of Benzodiazepine Prescription Practices in Elderly Appalachians with Dementia via the Appalachian Informatics Platform: Longitudinal Study %A Bhardwaj,Niharika %A Cecchetti,Alfred A %A Murughiyan,Usha %A Neitch,Shirley %+ Department of Clinical and Translational Science, Joan C Edwards School of Medicine, Marshall University, 1600 Medical Center Drive, Suite 265, Huntington, WV, 25701, United States, 1 304 691 5397, bhardwaj1@marshall.edu %K dementia %K Alzheimer disease %K benzodiazepines %K Appalachia %K geriatrics %K informatics platform %K interactive visualization %K eHealth %K clinical data %D 2020 %7 4.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: Caring for the growing dementia population with complex health care needs in West Virginia has been challenging due to its large, sizably rural-dwelling geriatric population and limited resource availability. Objective: This paper aims to illustrate the application of an informatics platform to drive dementia research and quality care through a preliminary study of benzodiazepine (BZD) prescription patterns and its effects on health care use by geriatric patients. Methods: The Maier Institute Data Mart, which contains clinical and billing data on patients aged 65 years and older (N=98,970) seen within our clinics and hospital, was created. Relevant variables were analyzed to identify BZD prescription patterns and calculate related charges and emergency department (ED) use. Results: Nearly one-third (4346/13,910, 31.24%) of patients with dementia received at least one BZD prescription, 20% more than those without dementia. More women than men received at least one BZD prescription. On average, patients with dementia and at least one BZD prescription sustained higher charges and visited the ED more often than those without one. Conclusions: The Appalachian Informatics Platform has the potential to enhance dementia care and research through a deeper understanding of dementia, data enrichment, risk identification, and care gap analysis. %M 32749226 %R 10.2196/18389 %U https://medinform.jmir.org/2020/8/e18389 %U https://doi.org/10.2196/18389 %U http://www.ncbi.nlm.nih.gov/pubmed/32749226 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e15932 %T Prediction of Cardiac Arrest in the Emergency Department Based on Machine Learning and Sequential Characteristics: Model Development and Retrospective Clinical Validation Study %A Hong,Sungjun %A Lee,Sungjoo %A Lee,Jeonghoon %A Cha,Won Chul %A Kim,Kyunga %+ Department of Emergency Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Irwon-ro, Gangnam-gu, Seoul, 06351, Republic of Korea, 82 10 5386 6597, wc.cha@samsung.com %K machine learning %K cardiac arrest prediction %K emergency department %K sequential characteristics %K clinical validity %D 2020 %7 4.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: The development and application of clinical prediction models using machine learning in clinical decision support systems is attracting increasing attention. Objective: The aims of this study were to develop a prediction model for cardiac arrest in the emergency department (ED) using machine learning and sequential characteristics and to validate its clinical usefulness. Methods: This retrospective study was conducted with ED patients at a tertiary academic hospital who suffered cardiac arrest. To resolve the class imbalance problem, sampling was performed using propensity score matching. The data set was chronologically allocated to a development cohort (years 2013 to 2016) and a validation cohort (year 2017). We trained three machine learning algorithms with repeated 10-fold cross-validation. Results: The main performance parameters were the area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC). The random forest algorithm (AUROC 0.97; AUPRC 0.86) outperformed the recurrent neural network (AUROC 0.95; AUPRC 0.82) and the logistic regression algorithm (AUROC 0.92; AUPRC=0.72). The performance of the model was maintained over time, with the AUROC remaining at least 80% across the monitored time points during the 24 hours before event occurrence. Conclusions: We developed a prediction model of cardiac arrest in the ED using machine learning and sequential characteristics. The model was validated for clinical usefulness by chronological visualization focused on clinical usability. %M 32749227 %R 10.2196/15932 %U http://medinform.jmir.org/2020/8/e15932/ %U https://doi.org/10.2196/15932 %U http://www.ncbi.nlm.nih.gov/pubmed/32749227 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e18089 %T Assessment of the Robustness of Convolutional Neural Networks in Labeling Noise by Using Chest X-Ray Images From Multiple Centers %A Jang,Ryoungwoo %A Kim,Namkug %A Jang,Miso %A Lee,Kyung Hwa %A Lee,Sang Min %A Lee,Kyung Hee %A Noh,Han Na %A Seo,Joon Beom %+ Department of Convergence Medicine, Asan Medical Center, University of Ulsan College of Medicine, 88 Olympic-Ro 43-Gil, Songpa-Gu, Seoul, Korea, Seoul, Republic of Korea, 82 10 3017 4282, namkugkim@gmail.com %K deep learning %K convolutional neural network %K NIH dataset %K CheXpert dataset %K robustness %D 2020 %7 4.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: Computer-aided diagnosis on chest x-ray images using deep learning is a widely studied modality in medicine. Many studies are based on public datasets, such as the National Institutes of Health (NIH) dataset and the Stanford CheXpert dataset. However, these datasets are preprocessed by classical natural language processing, which may cause a certain extent of label errors. Objective: This study aimed to investigate the robustness of deep convolutional neural networks (CNNs) for binary classification of posteroanterior chest x-ray through random incorrect labeling. Methods: We trained and validated the CNN architecture with different noise levels of labels in 3 datasets, namely, Asan Medical Center-Seoul National University Bundang Hospital (AMC-SNUBH), NIH, and CheXpert, and tested the models with each test set. Diseases of each chest x-ray in our dataset were confirmed by a thoracic radiologist using computed tomography (CT). Receiver operating characteristic (ROC) and area under the curve (AUC) were evaluated in each test. Randomly chosen chest x-rays of public datasets were evaluated by 3 physicians and 1 thoracic radiologist. Results: In comparison with the public datasets of NIH and CheXpert, where AUCs did not significantly drop to 16%, the AUC of the AMC-SNUBH dataset significantly decreased from 2% label noise. Evaluation of the public datasets by 3 physicians and 1 thoracic radiologist showed an accuracy of 65%-80%. Conclusions: The deep learning–based computer-aided diagnosis model is sensitive to label noise, and computer-aided diagnosis with inaccurate labels is not credible. Furthermore, open datasets such as NIH and CheXpert need to be distilled before being used for deep learning–based computer-aided diagnosis. %M 32749222 %R 10.2196/18089 %U https://medinform.jmir.org/2020/8/e18089 %U https://doi.org/10.2196/18089 %U http://www.ncbi.nlm.nih.gov/pubmed/32749222 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e16948 %T Semantic Deep Learning: Prior Knowledge and a Type of Four-Term Embedding Analogy to Acquire Treatments for Well-Known Diseases %A Arguello Casteleiro,Mercedes %A Des Diz,Julio %A Maroto,Nava %A Fernandez Prieto,Maria Jesus %A Peters,Simon %A Wroe,Chris %A Sevillano Torrado,Carlos %A Maseda Fernandez,Diego %A Stevens,Robert %+ Department of Computer Science, University of Manchester, Kilburn Building, Oxford Road, M13 9PL, Manchester, , United Kingdom, 44 161 275 6251, robert.stevens@manchester.ac.uk %K evidence-based practice %K artificial intelligence %K deep learning %K semantic deep learning %K analogical reasoning %K embedding analogies %K PubMed %D 2020 %7 6.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: How to treat a disease remains to be the most common type of clinical question. Obtaining evidence-based answers from biomedical literature is difficult. Analogical reasoning with embeddings from deep learning (embedding analogies) may extract such biomedical facts, although the state-of-the-art focuses on pair-based proportional (pairwise) analogies such as man:woman::king:queen (“queen = −man +king +woman”). Objective: This study aimed to systematically extract disease treatment statements with a Semantic Deep Learning (SemDeep) approach underpinned by prior knowledge and another type of 4-term analogy (other than pairwise). Methods: As preliminaries, we investigated Continuous Bag-of-Words (CBOW) embedding analogies in a common-English corpus with five lines of text and observed a type of 4-term analogy (not pairwise) applying the 3CosAdd formula and relating the semantic fields person and death: “dagger = −Romeo +die +died” (search query: −Romeo +die +died). Our SemDeep approach worked with pre-existing items of knowledge (what is known) to make inferences sanctioned by a 4-term analogy (search query −x +z1 +z2) from CBOW and Skip-gram embeddings created with a PubMed systematic reviews subset (PMSB dataset). Stage1: Knowledge acquisition. Obtaining a set of terms, candidate y, from embeddings using vector arithmetic. Some n-gram pairs from the cosine and validated with evidence (prior knowledge) are the input for the 3cosAdd, seeking a type of 4-term analogy relating the semantic fields disease and treatment. Stage 2: Knowledge organization. Identification of candidates sanctioned by the analogy belonging to the semantic field treatment and mapping these candidates to unified medical language system Metathesaurus concepts with MetaMap. A concept pair is a brief disease treatment statement (biomedical fact). Stage 3: Knowledge validation. An evidence-based evaluation followed by human validation of biomedical facts potentially useful for clinicians. Results: We obtained 5352 n-gram pairs from 446 search queries by applying the 3CosAdd. The microaveraging performance of MetaMap for candidate y belonging to the semantic field treatment was F-measure=80.00% (precision=77.00%, recall=83.25%). We developed an empirical heuristic with some predictive power for clinical winners, that is, search queries bringing candidate y with evidence of a therapeutic intent for target disease x. The search queries -asthma +inhaled_corticosteroids +inhaled_corticosteroid and -epilepsy +valproate +antiepileptic_drug were clinical winners, finding eight evidence-based beneficial treatments. Conclusions: Extracting treatments with therapeutic intent by analogical reasoning from embeddings (423K n-grams from the PMSB dataset) is an ambitious goal. Our SemDeep approach is knowledge-based, underpinned by embedding analogies that exploit prior knowledge. Biomedical facts from embedding analogies (4-term type, not pairwise) are potentially useful for clinicians. The heuristic offers a practical way to discover beneficial treatments for well-known diseases. Learning from deep learning models does not require a massive amount of data. Embedding analogies are not limited to pairwise analogies; hence, analogical reasoning with embeddings is underexploited. %M 32759099 %R 10.2196/16948 %U https://medinform.jmir.org/2020/8/e16948 %U https://doi.org/10.2196/16948 %U http://www.ncbi.nlm.nih.gov/pubmed/32759099 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e19892 %T Decompensation in Critical Care: Early Prediction of Acute Heart Failure Onset %A Essay,Patrick %A Balkan,Baran %A Subbian,Vignesh %+ College of Engineering, The University of Arizona, 1127 E James E Rogers Way, Tucson, AZ, 85721-0020, United States, 1 4024305524, p.essay@icloud.com %K critical care %K heart failure %K intensive care units %K machine learning %K time series %K heart %K cardiology %K prediction %K chronic disease %K ICU %K intensive care unit %D 2020 %7 7.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: Heart failure is a leading cause of mortality and morbidity worldwide. Acute heart failure, broadly defined as rapid onset of new or worsening signs and symptoms of heart failure, often requires hospitalization and admission to the intensive care unit (ICU). This acute condition is highly heterogeneous and less well-understood as compared to chronic heart failure. The ICU, through detailed and continuously monitored patient data, provides an opportunity to retrospectively analyze decompensation and heart failure to evaluate physiological states and patient outcomes. Objective: The goal of this study is to examine the prevalence of cardiovascular risk factors among those admitted to ICUs and to evaluate combinations of clinical features that are predictive of decompensation events, such as the onset of acute heart failure, using machine learning techniques. To accomplish this objective, we leveraged tele-ICU data from over 200 hospitals across the United States. Methods: We evaluated the feasibility of predicting decompensation soon after ICU admission for 26,534 patients admitted without a history of heart failure with specific heart failure risk factors (ie, coronary artery disease, hypertension, and myocardial infarction) and 96,350 patients admitted without risk factors using remotely monitored laboratory, vital signs, and discrete physiological measurements. Multivariate logistic regression and random forest models were applied to predict decompensation and highlight important features from combinations of model inputs from dissimilar data. Results: The most prevalent risk factor in our data set was hypertension, although most patients diagnosed with heart failure were admitted to the ICU without a risk factor. The highest heart failure prediction accuracy was 0.951, and the highest area under the receiver operating characteristic curve was 0.9503 with random forest and combined vital signs, laboratory values, and discrete physiological measurements. Random forest feature importance also highlighted combinations of several discrete physiological features and laboratory measures as most indicative of decompensation. Timeline analysis of aggregate vital signs revealed a point of diminishing returns where additional vital signs data did not continue to improve results. Conclusions: Heart failure risk factors are common in tele-ICU data, although most patients that are diagnosed with heart failure later in an ICU stay presented without risk factors making a prediction of decompensation critical. Decompensation was predicted with reasonable accuracy using tele-ICU data, and optimal data extraction for time series vital signs data was identified near a 200-minute window size. Overall, results suggest combinations of laboratory measurements and vital signs are viable for early and continuous prediction of patient decompensation. %M 32663162 %R 10.2196/19892 %U http://medinform.jmir.org/2020/8/e19892/ %U https://doi.org/10.2196/19892 %U http://www.ncbi.nlm.nih.gov/pubmed/32663162 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e18715 %T Machine Learning Model Based on Transthoracic Bioimpedance and Heart Rate Variability for Lung Fluid Accumulation Detection: Prospective Clinical Study %A Reljin,Natasa %A Posada-Quintero,Hugo F %A Eaton-Robb,Caitlin %A Binici,Sophia %A Ensom,Emily %A Ding,Eric %A Hayes,Anna %A Riistama,Jarno %A Darling,Chad %A McManus,David %A Chon,Ki H %+ Department of Biomedical Engineering, University of Connecticut, 67 N Eagleville Rd, Storrs, Mansfield, CT, 06269, United States, 1 5088739247, h.posada@uconn.edu %K heart failure %K transthoracic bioimpedance %K heart rate variability %K fluid accumulation %K autonomic nervous system %K machine learning %K cardiology %D 2020 %7 27.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: Accumulation of excess body fluid and autonomic dysregulation are clinically important characteristics of acute decompensated heart failure. We hypothesized that transthoracic bioimpedance, a noninvasive, simple method for measuring fluid retention in lungs, and heart rate variability, an assessment of autonomic function, can be used for detection of fluid accumulation in patients with acute decompensated heart failure. Objective: We aimed to evaluate the performance of transthoracic bioimpedance and heart rate variability parameters obtained using a fluid accumulation vest with carbon black–polydimethylsiloxane dry electrodes in a prospective clinical study (System for Heart Failure Identification Using an External Lung Fluid Device; SHIELD). Methods: We computed 15 parameters: 8 were calculated from the model to fit Cole-Cole plots from transthoracic bioimpedance measurements (extracellular, intracellular, intracellular-extracellular difference, and intracellular-extracellular parallel circuit resistances as well as fitting error, resonance frequency, tissue heterogeneity, and cellular membrane capacitance), and 7 were based on linear (mean heart rate, low-frequency components of heart rate variability, high-frequency components of heart rate variability, normalized low-frequency components of heart rate variability, normalized high-frequency components of heart rate variability) and nonlinear (principal dynamic mode index of sympathetic function, and principal dynamic mode index of parasympathetic function) analysis of heart rate variability. We compared the values of these parameters between 3 participant data sets: control (n=32, patients who did not have heart failure), baseline (n=23, patients with acute decompensated heart failure taken at the time of admittance to the hospital), and discharge (n=17, patients with acute decompensated heart failure taken at the time of discharge from hospital). We used several machine learning approaches to classify participants with fluid accumulation (baseline) and without fluid accumulation (control and discharge), termed with fluid and without fluid groups, respectively. Results: Among the 15 parameters, 3 transthoracic bioimpedance (extracellular resistance, R0; difference in extracellular-intracellular resistance, R0 – R∞, and tissue heterogeneity, α) and 3 heart rate variability (high-frequency, normalized low-frequency, and normalized high-frequency components) parameters were found to be the most discriminatory between groups (patients with and patients without heart failure). R0 and R0 – R∞ had significantly lower values for patients with heart failure than for those without heart failure (R0: P=.006; R0 – R∞: P=.001), indicating that a higher volume of fluids accumulated in the lungs of patients with heart failure. A cubic support vector machine model using the 5 parameters achieved an accuracy of 92% for with fluid and without fluid group classification. The transthoracic bioimpedance parameters were related to intra- and extracellular fluid, whereas the heart rate variability parameters were mostly related to sympathetic activation. Conclusions: This is useful, for instance, for an in-home diagnostic wearable to detect fluid accumulation. Results suggest that fluid accumulation, and subsequently acute decompensated heart failure detection, could be performed using transthoracic bioimpedance and heart rate variability measurements acquired with a wearable vest. %M 32852277 %R 10.2196/18715 %U http://medinform.jmir.org/2020/8/e18715/ %U https://doi.org/10.2196/18715 %U http://www.ncbi.nlm.nih.gov/pubmed/32852277 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e19870 %T Using Dual Neural Network Architecture to Detect the Risk of Dementia With Community Health Data: Algorithm Development and Validation Study %A Shen,Xiao %A Wang,Guanjin %A Kwan,Rick Yiu-Cho %A Choi,Kup-Sze %+ Centre for Smart Health, School of Nursing, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, 852 3400 3214, hskschoi@polyu.edu.hk %K cognitive screening %K dementia risk %K dual neural network %K predictive models %K primary care %D 2020 %7 31.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: Recent studies have revealed lifestyle behavioral risk factors that can be modified to reduce the risk of dementia. As modification of lifestyle takes time, early identification of people with high dementia risk is important for timely intervention and support. As cognitive impairment is a diagnostic criterion of dementia, cognitive assessment tools are used in primary care to screen for clinically unevaluated cases. Among them, Mini-Mental State Examination (MMSE) is a very common instrument. However, MMSE is a questionnaire that is administered when symptoms of memory decline have occurred. Early administration at the asymptomatic stage and repeated measurements would lead to a practice effect that degrades the effectiveness of MMSE when it is used at later stages. Objective: The aim of this study was to exploit machine learning techniques to assist health care professionals in detecting high-risk individuals by predicting the results of MMSE using elderly health data collected from community-based primary care services. Methods: A health data set of 2299 samples was adopted in the study. The input data were divided into two groups of different characteristics (ie, client profile data and health assessment data). The predictive output was the result of two-class classification of the normal and high-risk cases that were defined based on MMSE. A dual neural network (DNN) model was proposed to obtain the latent representations of the two groups of input data separately, which were then concatenated for the two-class classification. Mean and k-nearest neighbor were used separately to tackle missing data, whereas a cost-sensitive learning (CSL) algorithm was proposed to deal with class imbalance. The performance of the DNN was evaluated by comparing it with that of conventional machine learning methods. Results: A total of 16 predictive models were built using the elderly health data set. Among them, the proposed DNN with CSL outperformed in the detection of high-risk cases. The area under the receiver operating characteristic curve, average precision, sensitivity, and specificity reached 0.84, 0.88, 0.73, and 0.80, respectively. Conclusions: The proposed method has the potential to serve as a tool to screen for elderly people with cognitive impairment and predict high-risk cases of dementia at the asymptomatic stage, providing health care professionals with early signals that can prompt suggestions for a follow-up or a detailed diagnosis. %M 32865498 %R 10.2196/19870 %U https://medinform.jmir.org/2020/8/e19870 %U https://doi.org/10.2196/19870 %U http://www.ncbi.nlm.nih.gov/pubmed/32865498 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e23253 %T Correction: Undergraduate Medical Students’ Search for Health Information Online: Explanatory Cross-Sectional Study %A Loda,Teresa %A Erschens,Rebecca %A Junne,Florian %A Stengel,Andreas %A Zipfel,Stephan %A Herrmann-Werner,Anne %+ Medical Department VI/Psychosomatic Medicine and Psychotherapy, University Hospital Tuebingen, Osianderstr 5, Tuebingen, 72076, Germany, 49 07071 ext 2986719, rebecca.erschens@med.uni-tuebingen.de %D 2020 %7 11.8.2020 %9 Corrigenda and Addenda %J JMIR Med Inform %G English %X %M 32780713 %R 10.2196/23253 %U http://medinform.jmir.org/2020/8/e23253/ %U https://doi.org/10.2196/23253 %U http://www.ncbi.nlm.nih.gov/pubmed/32780713 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e18189 %T Artificial Intelligence for Caregivers of Persons With Alzheimer’s Disease and Related Dementias: Systematic Literature Review %A Xie,Bo %A Tao,Cui %A Li,Juan %A Hilsabeck,Robin C %A Aguirre,Alyssa %+ School of Nursing, The University of Texas at Austin, 1710 Red River, Austin, TX, 78712, United States, 1 512 232 5788, boxie@utexas.edu %K Alzheimer disease %K dementia %K caregiving %K technology %K artificial intelligence %D 2020 %7 20.8.2020 %9 Review %J JMIR Med Inform %G English %X Background: Artificial intelligence (AI) has great potential for improving the care of persons with Alzheimer’s disease and related dementias (ADRD) and the quality of life of their family caregivers. To date, however, systematic review of the literature on the impact of AI on ADRD management has been lacking. Objective: This paper aims to (1) identify and examine literature on AI that provides information to facilitate ADRD management by caregivers of individuals diagnosed with ADRD and (2) identify gaps in the literature that suggest future directions for research. Methods: Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for conducting systematic literature reviews, during August and September 2019, we performed 3 rounds of selection. First, we searched predetermined keywords in PubMed, Cumulative Index to Nursing and Allied Health Literature Plus with Full Text, PsycINFO, IEEE Xplore Digital Library, and the ACM Digital Library. This step generated 113 nonduplicate results. Next, we screened the titles and abstracts of the 113 papers according to inclusion and exclusion criteria, after which 52 papers were excluded and 61 remained. Finally, we screened the full text of the remaining papers to ensure that they met the inclusion or exclusion criteria; 31 papers were excluded, leaving a final sample of 30 papers for analysis. Results: Of the 30 papers, 20 reported studies that focused on using AI to assist in activities of daily living. A limited number of specific daily activities were targeted. The studies’ aims suggested three major purposes: (1) to test the feasibility, usability, or perceptions of prototype AI technology; (2) to generate preliminary data on the technology’s performance (primarily accuracy in detecting target events, such as falls); and (3) to understand user needs and preferences for the design and functionality of to-be-developed technology. The majority of the studies were qualitative, with interviews, focus groups, and observation being their most common methods. Cross-sectional surveys were also common, but with small convenience samples. Sample sizes ranged from 6 to 106, with the vast majority on the low end. The majority of the studies were descriptive, exploratory, and lacking theoretical guidance. Many studies reported positive outcomes in favor of their AI technology’s feasibility and satisfaction; some studies reported mixed results on these measures. Performance of the technology varied widely across tasks. Conclusions: These findings call for more systematic designs and evaluations of the feasibility and efficacy of AI-based interventions for caregivers of people with ADRD. These gaps in the research would be best addressed through interdisciplinary collaboration, incorporating complementary expertise from the health sciences and computer science/engineering–related fields. %M 32663146 %R 10.2196/18189 %U http://medinform.jmir.org/2020/8/e18189/ %U https://doi.org/10.2196/18189 %U http://www.ncbi.nlm.nih.gov/pubmed/32663146 %0 Journal Article %@ 2291-9694 %I JMIR Publications %V 8 %N 8 %P e20992 %T Nationwide Results of COVID-19 Contact Tracing in South Korea: Individual Participant Data From an Epidemiological Survey %A Lee,Seung Won %A Yuh,Woon Tak %A Yang,Jee Myung %A Cho,Yoon-Sik %A Yoo,In Kyung %A Koh,Hyun Yong %A Marshall,Dominic %A Oh,Donghwan %A Ha,Eun Kyo %A Han,Man Yong %A Yon,Dong Keon %+ Armed Force Medical Command, Republic of Korea Armed Forces, 81 Saemaeul-ro 177, Seongnam, 463-040, Republic of Korea, 82 2 6935 2476, yonkkang@gmail.com %K COVID-19 %K contact tracing %K coronavirus %K South Korea %K survey %K health data %K epidemiology %K transmission %D 2020 %7 25.8.2020 %9 Original Paper %J JMIR Med Inform %G English %X Background: Evidence regarding the effectiveness of contact tracing of COVID-19 and the related social distancing is limited and inconclusive. Objective: This study aims to investigate the epidemiological characteristics of SARS-CoV-2 transmission in South Korea and evaluate whether a social distancing campaign is effective in mitigating the spread of COVID-19. Methods: We used contract tracing data to investigate the epidemic characteristics of SARS-CoV-2 transmission in South Korea and evaluate whether a social distancing campaign was effective in mitigating the spread of COVID-19. We calculated the mortality rate for COVID-19 by infection type (cluster vs noncluster) and tested whether new confirmed COVID-19 trends changed after a social distancing campaign. Results: There were 2537 patients with confirmed COVID-19 who completed the epidemiologic survey: 1305 (51.4%) cluster cases and 1232 (48.6%) noncluster cases. The mortality rate was significantly higher in cluster cases linked to medical facilities (11/143, 7.70% vs 5/1232, 0.41%; adjusted percentage difference 7.99%; 95% CI 5.83 to 10.14) and long-term care facilities (19/221, 8.60% vs 5/1232, 0.41%; adjusted percentage difference 7.56%; 95% CI 5.66 to 9.47) than in noncluster cases. The change in trends of newly confirmed COVID-19 cases before and after the social distancing campaign was significantly negative in the entire cohort (adjusted trend difference –2.28; 95% CI –3.88 to –0.68) and the cluster infection group (adjusted trend difference –0.96; 95% CI –1.83 to –0.09). Conclusions: In a nationwide contact tracing study in South Korea, COVID-19 linked to medical and long-term care facilities significantly increased the risk of mortality compared to noncluster COVID-19. A social distancing campaign decreased the spread of COVID-19 in South Korea and differentially affected cluster infections of SARS-CoV-2. %M 32784189 %R 10.2196/20992 %U http://medinform.jmir.org/2020/8/e20992/ %U https://doi.org/10.2196/20992 %U http://www.ncbi.nlm.nih.gov/pubmed/32784189