PLOS digital health最新文献

Meta-researchers commonly leverage tools that infer gender from first names, especially when studying gender disparities. However, tools vary in their accuracy, ease of use, and cost. The objective of this study was to compare the accuracy and cost of the commercial software Genderize and Gender API, and the open-source gender R package. Differences in binary gender prediction accuracy between the three services were evaluated. Gender prediction accuracy was tested on a multi-national dataset of 32,968 gender-labeled clinical trial authors. Additionally, two datasets from previous studies with 5779 and 6131 names, respectively, were re-evaluated with modern implementations of Genderize and Gender API. The gender inference accuracy of Genderize and Gender API were compared, both with and without supplying trialists' country of origin in the API call. The accuracy of the gender R package was only evaluated without supplying countries of origin. The accuracy of Genderize, Gender API, and the gender R package were defined as the percentage of correct gender predictions. Accuracy differences between methods were evaluated using McNemar's test. Genderize and Gender API demonstrated 96.6% and 96.1% accuracy, respectively, when countries of origin were not supplied in the API calls. Genderize and Gender API achieved the highest accuracy when predicting the gender of German authors with accuracies greater than 98%. Genderize and Gender API were least accurate with South Korean, Chinese, Singaporean, and Taiwanese authors, demonstrating below 82% accuracy. Genderize can provide similar accuracy to Gender API while being 4.85x less expensive. The gender R package achieved below 86% accuracy on the full dataset. In the replication studies, Genderize and gender API demonstrated better performance than in the original publications. Our results indicate that Genderize and Gender API achieve similar accuracy on a multinational dataset. The gender R package is uniformly less accurate than Genderize and Gender API.

Aims: Patients visiting the emergency department (ED) or hospitalized for heart failure (HF) are at increased risk for subsequent adverse outcomes, however effective risk stratification remains challenging. We utilized a machine-learning (ML)-based approach to identify HF patients at risk of adverse outcomes after an ED visit or hospitalization using a large regional administrative healthcare data system.

Methods and results: Patients visiting the ED or hospitalized with HF between 2002-2016 in Alberta, Canada were included. Outcomes of interest were 30-day and 1-year HF-related ED visits, HF hospital readmission or all-cause mortality. We applied a feature extraction method using deep feature synthesis from multiple sources of health data and compared performance of a gradient boosting algorithm (CatBoost) with logistic regression modelling. The area under receiver operating characteristic curve (AUC-ROC) was used to assess model performance. We included 50,630 patients with 93,552 HF ED visits/hospitalizations. At 30-day follow-up in the holdout validation cohort, the AUC-ROC for the combined endpoint of HF ED visit, HF hospital readmission or death for the Catboost and logistic regression models was 74.16 (73.18-75.11) versus 62.25 (61.25-63.18), respectively. At 1-year follow-up corresponding values were 76.80 (76.1-77.47) versus 69.52 (68.77-70.26), respectively. AUC-ROC values for the endpoint of all-cause death alone at 30-days and 1-year follow-up were 83.21 (81.83-84.41) versus 69.53 (67.98-71.18), and 85.73 (85.14-86.29) versus 69.40 (68.57-70.26), for the CatBoost and logistic regression models, respectively.

Conclusions: ML-based modelling with deep feature synthesis provided superior risk stratification for HF patients at 30-days and 1-year follow-up after an ED visit or hospitalization using data from a large administrative regional healthcare system.

Health information on the Internet has a ubiquitous influence on health consumers' behaviour. Searching and evaluating online health information poses a real challenge for many health consumers. To our knowledge, our systematic review paper is the first to explore the interventions targeting lay people to improve their e-health literacy skills. Our paper aims to explore interventions to improve laypeople ability to identify trustworthy online health information. The search was conducted on Ovid Medline, Embase, Cochrane database, Academic Search Complete, and APA psych info. Publications were selected by screening title, abstract, and full text, then manual review of reference lists of selected publications. Data was extracted from eligible studies on an excel sheet about the types of interventions, the outcomes of the interventions and whether they are effective, and the barriers and facilitators for using the interventions by consumers. A mixed-methods appraisal tool was used to appraise evidence from quantitative, qualitative, and mixed-methods studies. Whittemore and Knafl's integrative review approach was used as a guidance for narrative synthesis. The total number of included studies is twelve. Media literacy interventions are the most common type of interventions. Few studies measured the effect of the interventions on patient health outcomes. All the procedural and navigation/ evaluation skills-building interventions are significantly effective. Computer/internet illiteracy and the absence of guidance/facilitators are significant barriers to web-based intervention use. Few interventions are distinguished by its implementation in a context tailored to consumers, using a human-centred design approach, and delivery through multiple health stakeholders' partnership. There is potential for further research to understand how to improve consumers health information use focusing on collaborative learning, using human-centred approaches, and addressing the social determinants of health.

Innovative information-sharing techniques and rapid access to stored research data as scientific currency have proved highly beneficial in healthcare and health research. Yet, researchers often experience conflict between data sharing to promote health-related scientific knowledge for the common good and their personal academic advancement. There is a scarcity of studies exploring the perspectives of health researchers in sub-Saharan Africa (SSA) regarding the challenges with data sharing in the context of data-intensive research. The study began with a quantitative survey and research, after which the researchers engaged in a qualitative study. This qualitative cross-sectional baseline study reports on the challenges faced by health researchers, in terms of data sharing. In-depth interviews were conducted via Microsoft Teams between July 2022 and April 2023 with 16 health researchers from 16 different countries across SSA. We employed purposive and snowballing sampling techniques to invite participants via email. The recorded interviews were transcribed, coded and analysed thematically using ATLAS.ti. Five recurrent themes and several subthemes emerged related to (1) individual researcher concerns (fears regarding data sharing, publication and manuscript pressure), (2) structural issues impacting data sharing, (3) recognition in academia (scooping of research data, acknowledgement and research incentives) (4) ethical challenges experienced by health researchers in SSA (confidentiality and informed consent, commercialisation and benefit sharing) and (5) legal lacunae (gaps in laws and regulations). Significant discomfort about data sharing exists amongst health researchers in this sample of respondents from SSA, resulting in a reluctance to share data despite acknowledging the scientific benefits of such sharing. This discomfort is related to the lack of adequate guidelines and governance processes in the context of health research collaborations, both locally and internationally. Consequently, concerns about ethical and legal issues are increasing. Resources are needed in SSA to improve the quality, value and veracity of data-as these are ethical imperatives. Strengthening data governance via robust guidelines, legislation and appropriate data sharing agreements will increase trust amongst health researchers and data donors alike.

Identifying the onset of patient deterioration is challenging despite the potential to respond to patients earlier with better vital sign monitoring and rapid response team (RRT) activation. In this study an ECG based software as a medical device, the Analytic for Hemodynamic Instability Predictive Index (AHI-PI), was compared to the vital signs of heart rate, blood pressure, and respiratory rate, evaluating how early it indicated risk before an RRT activation. A higher proportion of the events had risk indication by AHI-PI (92.71%) than by vital signs (41.67%). AHI-PI indicated risk early, with an average of over a day before RRT events. In events whose risks were indicated by both AHI-PI and vital signs, AHI-PI demonstrated earlier recognition of deterioration compared to vital signs. A case-control study showed that situations requiring RRTs were more likely to have AHI-PI risk indication than those that did not. The study derived several insights in support of AHI-PI's efficacy as a clinical decision support system. The findings demonstrated AHI-PI's potential to serve as a reliable predictor of future RRT events. It could potentially help clinicians recognize early clinical deterioration and respond to those unnoticed by vital signs, thereby helping clinicians improve clinical outcomes.

Electronic Health Records (EHRs) are increasingly used to develop machine learning models in predictive medicine. There has been limited research on utilizing machine learning methods to predict childhood obesity and related disparities in classifier performance among vulnerable patient subpopulations. In this work, classification models are developed to recognize pediatric obesity using temporal condition patterns obtained from patient EHR data in a U.S. study population. We trained four machine learning algorithms (Logistic Regression, Random Forest, Gradient Boosted Trees, and Neural Networks) to classify cases and controls as obesity positive or negative, and optimized hyperparameter settings through a bootstrapping methodology. To assess the classifiers for bias, we studied model performance by population subgroups then used permutation analysis to identify the most predictive features for each model and the demographic characteristics of patients with these features. Mean AUC-ROC values were consistent across classifiers, ranging from 0.72-0.80. Some evidence of bias was identified, although this was through the models performing better for minority subgroups (African Americans and patients enrolled in Medicaid). Permutation analysis revealed that patients from vulnerable population subgroups were over-represented among patients with the most predictive diagnostic patterns. We hypothesize that our models performed better on under-represented groups because the features more strongly associated with obesity were more commonly observed among minority patients. These findings highlight the complex ways that bias may arise in machine learning models and can be incorporated into future research to develop a thorough analytical approach to identify and mitigate bias that may arise from features and within EHR datasets when developing more equitable models.

Background: Investigators often use claims data to estimate the diagnosis timing of chronic conditions. However, misclassification of chronic conditions is common due to variability in healthcare utilization and in claims history across patients.

Objective: We aimed to quantify the effect of various Medicare fee-for-service continuous enrollment period and lookback period (LBP) on misclassification of COPD and sample size.

Methods: A stepwise tutorial to classify COPD, based on its diagnosis timing relative to lung cancer diagnosis using the Surveillance Epidemiology and End Results cancer registry linked to Medicare insurance claims. We used 3 approaches varying the LBP and required continuous enrollment (i.e., observability) period between 1 to 5 years. Patients with lung cancer were classified based on their COPD related healthcare utilization into 3 groups: pre-existing COPD (diagnosis at least 3 months before lung cancer diagnosis), concurrent COPD (diagnosis during the -/+ 3months of lung cancer diagnosis), and non-COPD. Among those with 5 years of continuous enrollment, we estimated the sensitivity of the LBP to ascertain COPD diagnosis as the number of patients with pre-existing COPD using a shorter LBP divided by the number of patients with pre-existing COPD using a longer LBP.

Results: Extending the LBP from 1 to 5 years increased prevalence of pre-existing COPD from ~ 36% to 51%, decreased both concurrent COPD from ~ 34% to 23% and non-COPD from ~ 29% to 25%. There was minimal effect of extending the required continuous enrollment period beyond one year across various LBPs. In those with 5 years of continuous enrollment, sensitivity of COPD classification (95% CI) increased with longer LBP from 70.1% (69.7% to 70.4%) for one-year LBP to 100% for 5-years LBP.

Conclusion: The length of optimum LBP and continuous enrollment period depends on the context of the research question and the data generating mechanisms. Among Medicare beneficiaries, the best approach to identify diagnosis timing of COPD relative to lung cancer diagnosis is to use all available LBP with at least one year of required continuous enrollment.

Recent work in machine learning for healthcare has raised concerns about patient privacy and algorithmic fairness. Previous work has shown that self-reported race can be predicted from medical data that does not explicitly contain racial information. However, the extent of data identification is unknown, and we lack ways to develop models whose outcomes are minimally affected by such information. Here we systematically investigated the ability of time-series electronic health record data to predict patient static information. We found that not only the raw time-series data, but also learned representations from machine learning models, can be trained to predict a variety of static information with area under the receiver operating characteristic curve as high as 0.851 for biological sex, 0.869 for binarized age and 0.810 for self-reported race. Such high predictive performance can be extended to various comorbidity factors and exists even when the model was trained for different tasks, using different cohorts, using different model architectures and databases. Given the privacy and fairness concerns these findings pose, we develop a variational autoencoder-based approach that learns a structured latent space to disentangle patient-sensitive attributes from time-series data. Our work thoroughly investigates the ability of machine learning models to encode patient static information from time-series electronic health records and introduces a general approach to protect patient-sensitive information for downstream tasks.

Objectives: To develop and validate a model to predict time-to-LTC admissions among individuals with dementia.

Design: Population-based retrospective cohort study using health administrative data.

Setting and participants: Community-dwelling older adults (65+) in Ontario living with dementia and assessed with the Resident Assessment Instrument for Home Care (RAI-HC) between April 1, 2010 and March 31, 2017.

Methods: Individuals in the derivation cohort (n = 95,813; assessed before March 31, 2015) were followed for up to 360 days after the index RAI-HC assessment for admission into LTC. We used a multivariable Fine Gray sub-distribution hazard model to predict the cumulative incidence of LTC entry while accounting for all-cause mortality as a competing risk. The model was validated in 34,038 older adults with dementia with an index RAI-HC assessment between April 1, 2015 and March 31, 2017.

Results: Within one year of a RAI-HC assessment, 35,513 (37.1%) individuals in the derivation cohort and 10,735 (31.5%) in the validation cohort entered LTC. Our algorithm was well-calibrated (Emax = 0.119, ICIavg = 0.057) and achieved a c-statistic of 0.707 (95% confidence interval: 0.703-0.712) in the validation cohort.

Conclusions and implications: We developed an algorithm to predict time to LTC entry among individuals living with dementia. This tool can inform care planning for individuals with dementia and their family caregivers.

Nosocomial infections and Antimicrobial Resistance (AMR) stand as formidable healthcare challenges on a global scale. To address these issues, various infection control protocols and personalized treatment strategies, guided by laboratory tests, aim to detect bloodstream infections (BSI) and assess the potential for AMR. In this study, we introduce a machine learning (ML) approach based on Multi-Objective Symbolic Regression (MOSR), an evolutionary approach to create ML models in the form of readable mathematical equations in a multi-objective way to overcome the limitation of standard single-objective approaches. This method leverages readily available clinical data collected upon admission to intensive care units, with the goal of predicting the presence of BSI and AMR. We further assess its performance by comparing it to established ML algorithms using both naturally imbalanced real-world data and data that has been balanced through oversampling techniques. Our findings reveal that traditional ML models exhibit subpar performance across all training scenarios. In contrast, MOSR, specifically configured to minimize false negatives by optimizing also for the F1-Score, outperforms other ML algorithms and consistently delivers reliable results, irrespective of the training set balance with F1-Score.22 and.28 higher than any other alternative. This research signifies a promising path forward in enhancing Antimicrobial Stewardship (AMS) strategies. Notably, the MOSR approach can be readily implemented on a large scale, offering a new ML tool to find solutions to these critical healthcare issues affected by limited data availability.