Using indication embeddings to represent patient health for drug safety studies.
Ontology highlight
ABSTRACT: Objective
The electronic health record is a rising resource for quantifying medical practice, discovering the adverse effects of drugs, and studying comparative effectiveness. One of the challenges of applying these methods to health care data is the high dimensionality of the health record. Methods to discover the effects of drugs in health data must account for tens of thousands of potentially relevant confounders. Our goal in this work is to reduce the dimensionality of the health data with the aim of accelerating the application of retrospective cohort studies to this data.Materials and methods
Here, we develop indication embeddings, a way to reduce the dimensionality of health data while capturing information relevant to treatment decisions. We evaluate these embeddings using external data on drug indications. Then, we use the embeddings as a substitute for medical history to match patients and develop evaluation metrics for these matches.Results
We demonstrate that these embeddings recover the therapeutic uses of drugs. We use embeddings as an informative representation of relationships between drugs, between health history events and drug prescriptions, and between patients at a particular time in their health history. We show that using embeddings to match cohorts improves the balance of the cohorts, even in terms of poorly measured risk factors like smoking.Discussion and conclusion
Unlike other embeddings inspired by word2vec, indication embeddings are specifically designed to capture the medical history leading to the prescription of a new drug. For retrospective cohort studies, our low-dimensional representation helps in finding comparator drugs and constructing comparator cohorts.
SUBMITTER: Melamed RD
PROVIDER: S-EPMC7751136 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Various patient safety interventions have been implemented since the late 1990s, but their evaluation has been lacking. To obtain basic information for prioritizing patient safety interventions, this study aimed to extract high-priority interventions in Japan and to identify the factors that influence the setting of priority. Six perspectives (contribution, dissemination, impact, cost, urgency, and priority) on 42 patient safety interventions classified into 3 levels (system, organizational, and clinical) were evaluated by Japanese experts using the Delphi technique. We examined the relationships of the levels and the perspectives on interventions with the transition of the consensus state in rounds 1 and 3. After extracting the high-priority interventions, a chi-squared test was used to examine the relationship of the levels and the impact/cost ratio with high priority. Regression models were used to examine the influence of each perspective on priority. There was a significant relationship between the level of interventions and the transition of the consensus state (p = 0.033). System-level interventions had a low probability of achieving consensus. "Human resources interventions," "professional education and training," "medication management/reconciliation protocols," "pay-for performance (P4P) schemes and financing for safety," "digital technology solutions to improve safety," and "hand hygiene initiatives" were extracted as high-priority interventions. The level and the impact/cost ratio of interventions had no significant relationships with high priority. In the regression model, dissemination and impact had an influence on priority (β = -0.628 and 0.941, respectively; adjusted R-squared = 0.646). The influence of impact and dissemination on the priority of interventions suggests that it is important to examine the dissemination degree and impact of interventions in each country for prioritizing interventions.
| S-EPMC7497979 | biostudies-literature
Project description:BackgroundWith the knowledge that patient safety incidents can significantly impact patients, providers and health-care organizations, greater emphasis on patient involvement as a means to mitigate risks warrants further research.ObjectiveTo understand whether patient perceptions of patient safety play a role in patient involvement in factual and challenging patient safety practices and whether the constructs of the Health Belief Model (HBM) help to explain such perceptions.DesignPartial least squares (PLS) analysis of survey data.Setting and participantsFour inpatient units located in two tertiary hospitals in Atlantic Canada. Patients discharged from participating units between November 2010 and January 2011.InterventionNone.ResultsA total of 217 of the 587 patient surveys were returned for a final response rate of 37.0%. The PLS analysis revealed relationships between patient perceptions of threat and self-efficacy and the performance of factual and challenging patient safety practices, explaining 46 and 42% of the variance, respectively.DiscussionThe results from this study provide evidence for the constructs and relationships set forth by the HBM. Perceptions of patient safety were shown to influence patient likelihood for engaging in selected patient safety practices. While perceptions of barriers and benefits and threats were found to be a contributing factor to patient involvement in patient safety practices, self-efficacy plays an important role as a mediating factor.ConclusionsOverall, the use of the HBM within patient safety provides for increased understanding of how such perceptions can be influenced to improve patient engagement in promoting safer health care.
| S-EPMC5810747 | biostudies-literature
Project description:BackgroundElectronic health records (EHRs) provide rich data on many domains not routinely available in other data, as such, they are a promising source to study changes in health outcomes using longitudinal study designs (eg, cohort studies, natural experiments, etc.). Yet, patient attrition rates in these data are unknown.ObjectiveThe objective of this study was to estimate overall and among adults with diabetes or hypertension: (1) patient attrition over a 3-year period at community health centers; and (2) the likelihood that patients with Medicaid permanently switched their source of primary care.Research designA retrospective cohort study of 2012-2017 data from the Accelerating Data Value Across a National Community Health Center Network (ADVANCE) Clinical Data Research Network of community health centers were used to assess EHR data attrition. Oregon Medicaid enrollment and claims data were used to estimate the likelihood of changing the source of primary care.SubjectsA total of 827,657 patients aged 19-64 with ≥1 ambulatory visit from 76 community health center systems across 20 states. In all, 232,891 Oregon Medicaid enrollees (aged 19-64) with a gap of ≥6 months following a claim for a visit billed to a primary care source.MeasuresPercentage of patients not returning within 3 years of their qualifying visit (attrition). The probability that a patient with Medicaid permanently changed their primary care source.ResultsAttrition over the 3 years averaged 33.5%; attrition rates were lower (<25%) among patients with diabetes and/or hypertension. Among Medicaid enrollees, the percentage of provider change after a 6-month gap between visits was 12% for community health center patients compared with 39% for single-provider practice patients. Over 3 years, the likelihood of a patient changing to a new provider increased with length of time since their last visit but remained lowest among community health center patients.ConclusionThis study demonstrates the use of the EHR dataset is a reliable source of data to support longitudinal studies while highlighting variability in attrition by primary care source and chronic conditions.
| S-EPMC7365658 | biostudies-literature
Project description:Drug labeling contains an 'INDICATIONS AND USAGE' that provides vital information to support clinical decision making and regulatory management. Effective extraction of drug indication information from free-text based resources could facilitate drug repositioning projects and help collect real-world evidence in support of secondary use of approved medicines. To enable AI-powered language models for the extraction of drug indication information, we used manual reading and curation to develop a Drug Indication Classification and Encyclopedia (DICE) based on FDA approved human prescription drug labeling. A DICE scheme with 7,231 sentences categorized into five classes (indications, contradictions, side effects, usage instructions, and clinical observations) was developed. To further elucidate the utility of the DICE, we developed nine different AI-based classifiers for the prediction of indications based on the developed DICE to comprehensively assess their performance. We found that the transformer-based language models yielded an average MCC of 0.887, outperforming the word embedding-based Bidirectional long short-term memory (BiLSTM) models (0.862) with a 2.82% improvement on the test set. The best classifiers were also used to extract drug indication information in DrugBank and achieved a high enrichment rate (>0.930) for this task. We found that domain-specific training could provide more explainable models without performance sacrifices and better generalization for external validation datasets. Altogether, the proposed DICE could be a standard resource for the development and evaluation of task-specific AI-powered, natural language processing (NLP) models.
| S-EPMC8366025 | biostudies-literature
Project description:Electronic health data are routinely used for population drug studies. Due to the ethical dilemma in carrying out experimental drug studies on pregnant women, the effects of medication usage during pregnancy on fetal and maternal outcomes are largely evaluated using this data collection medium. One major limitation in this type of study is the delayed inclusion of pregnancies in the cohort. For example, in the province of Quebec, Canada, a major pregnancy cohort only captured pregnancies after 20 weeks gestation. The purpose of this study was to demonstrate three methods that can be used to assess the extent of selection bias due to the delayed inclusion of pregnancies. We use causal directed acyclic graphs to explain the source of this selection bias. In an example involving a cohort of pregnant asthmatic women reconstructed from the linkage of administrative health databases from the province of Quebec, we use numerical derivations, a simulation study and a sensitivity analysis to investigate the potential for bias and loss of power due to the delayed inclusion. We find that this selection bias can be partially mitigated by controlling for variables related to (spontaneous or therapeutic) abortion and the outcome of interest. The three proposed methods allow for the pre and post hoc ascertainment of the bias. While delayed pregnancy inclusion selection bias (which includes "live birth bias") can produce substantial bias in pregnancy drug studies, all three methods are effective at producing estimates of the size of the bias.
| S-EPMC6149369 | biostudies-literature
Project description:Study objectivesTo assess the patient safety culture in Primary Health Care (PHC) setting after the transition to the Family Health Strategy (FHS) model in a Brazilian metropolitan area and compare the results between the categories of health care professionals.MethodsA cross-sectional study including 246 workers from primary health care services in Federal District, Brazil. Data collection took place from October to December 2019 through the Medical Office Survey on Patient Safety Culture (MOSPSC) application. Patient safety culture was considered positive when the score was above 60%. For comparisons between the categories of health care professionals', the ANOVA and Kruskal-Wallis test were used for composite percent positive scores, and Pearson's chi-square or Fishers exact test for frequency and percentage of positive responses.ResultsThe overall MOSPSC composite percent positive score was 49.9%. Among the 12 dimensions, only three showed a positive patient safety culture: Teamwork (73.1%), Organizational learning (62.9%), and Patient care tracking/follow-up (60.1%). The percentage of positive responses on overall quality assessment (78.1%) and overall patient safety assessment (78.0%) showed a positive evaluation. There was no significant difference in the composite percent positive score of overall MOSPSC (p = 0.135) and the percentage of positive responses on overall patient safety assessment (p = 0.156) between the categories of health care professionals. Overall quality assessment showed a significant difference between job roles (p < 0.001), in which nursing /health care technicians showed a significantly lower score than other job roles.ConclusionThe patient safety culture assessment showed a weakness in the patient safety in the PHC services. The MOSPSC and nine of its dimensions presented a negative safety culture assessment, regardless of the high scores in the overall patient safety and quality assessments.
| S-EPMC9321378 | biostudies-literature
Project description:BackgroundPatient safety in primary care is an emerging field of research with a growing evidence base in western countries but little has been explored in the Gulf Cooperation Council Countries (GCC) including the Sultanate of Oman. This study aimed to review the literature on the safety culture and patient safety measures used globally to inform the development of safety culture among health care workers in primary care with a particular focus on the Middle East.MethodsA systematic review of the literature. Searches were undertaken using Medline, EMBASE, CINAHL and Scopus from the year 2000 to 2014. Terms defining safety culture were combined with terms identifying patient safety and primary care.ResultsThe database searches identified 3072 papers that were screened for inclusion in the review. After the screening and verification, data were extracted from 28 papers that described safety culture in primary care. The global distribution of the articles is as follows: the Netherlands (7), the United States (5), Germany (4), the United Kingdom (1), Australia, Canada and Brazil (two for each country), and with one each from Turkey, Iran, Saudi Arabia and Kuwait. The characteristics of the included studies were grouped under the following themes: safety culture in primary care, incident reporting, safety climate and adverse events. The most common theme from 2011 onwards was the assessment of safety culture in primary care (13 studies, 46%). The most commonly used safety culture assessment tool is the Hospital survey on patient safety culture (HSOPSC) which has been used in developing countries in the Middle East.ConclusionsThis systematic review reveals that the most important first step is the assessment of safety culture in primary care which will provide a basic understanding to safety-related perceptions of health care providers. The HSOPSC has been commonly used in Kuwait, Turkey, and Iran.
| S-EPMC6026504 | biostudies-literature
Project description:Relatively few neonatal drug development studies have been conducted, but an increase is expected with the enactment of the Food and Drug Administration Safety and Innovation Act (FDASIA). Understanding the safety of drugs studied in neonates is complicated by the unique nature of the population and the level of illness. The objective of this study was to examine neonatal safety data submitted to the FDA in studies pursuant to the Best Pharmaceuticals for Children Act (BPCA) and the Pediatric Research Equity Act (PREA) between 1998 and 2015.FDA databases were searched for BPCA and/or PREA studies that enrolled neonates. Studies that enrolled a minimum of 3 neonates were analyzed for the presence and content of neonatal safety data.The analysis identified 40 drugs that were studied in 3 or more neonates. Of the 40 drugs, 36 drugs received a pediatric labeling change as a result of studies between 1998 and 2015, that included information from studies including neonates. Fourteen drugs were approved for use in neonates. Clinical trials for 20 of the drugs reported serious adverse events (SAEs) in neonates. The SAEs primarily involved cardiovascular events such as bradycardia and/or hypotension or laboratory abnormalities such as anemia, neutropenia, and electrolyte disturbances. Deaths were reported during studies of 9 drugs.Our analysis revealed that SAEs were reported in studies involving 20 of the 40 drugs evaluated in neonates, with deaths identified in 9 of those studies. Patients enrolled in studies were often critically ill, which complicated determination of whether an adverse event was drug-related. We conclude that the traditional means for collecting safety information in drug development trials needs to be adjusted for neonates and will require the collaboration of regulators, industry, and the clinical and research communities to establish appropriate definitions and reporting strategies for the neonatal population.
| S-EPMC5549935 | biostudies-other
Project description:The rapid expansion of protein sequence and structure databases has resulted in a significant number of proteins with ambiguous or unknown function. While advances in machine learning techniques hold great potential to fill this annotation gap, current methods for function prediction are unable to associate global function reliably to the specific residues responsible for that function. We address this issue by introducing PARSE (Protein Annotation by Residue-Specific Enrichment), a knowledge-based method which combines pre-trained embeddings of local structural environments with traditional statistical techniques to identify enriched functions with residue-level explainability. For the task of predicting the catalytic function of enzymes, PARSE achieves comparable or superior global performance to state-of-the-art machine learning methods (F1 score > 85%) while simultaneously annotating the specific residues involved in each function with much greater precision. Since it does not require supervised training, our method can make one-shot predictions for very rare functions and is not limited to a particular type of functional label (e.g. Enzyme Commission numbers or Gene Ontology codes). Finally, we leverage the AlphaFold Structure Database to perform functional annotation at a proteome scale. By applying PARSE to the dark proteome-predicted structures which cannot be classified into known structural families-we predict several novel bacterial metalloproteases. Each of these proteins shares a strongly conserved catalytic site despite highly divergent sequences and global folds, illustrating the value of local structure representations for new function discovery.
| S-EPMC10614799 | biostudies-literature
Project description:MotivationThe completion of the genome has paved the way for genome-wide association studies (GWAS), which explained certain proportions of heritability. GWAS are not optimally suited to detect non-linear effects in disease risk, possibly hidden in non-additive interactions (epistasis). Alternative methods for epistasis detection using, e.g. deep neural networks (DNNs) are currently under active development. However, DNNs are constrained by finite computational resources, which can be rapidly depleted due to increasing complexity with the sheer size of the genome. Besides, the curse of dimensionality complicates the task of capturing meaningful genetic patterns for DNNs; therefore necessitates dimensionality reduction.ResultsWe propose a method to compress single nucleotide polymorphism (SNP) data, while leveraging the linkage disequilibrium (LD) structure and preserving potential epistasis. This method involves clustering correlated SNPs into haplotype blocks and training per-block autoencoders to learn a compressed representation of the block's genetic content. We provide an adjustable autoencoder design to accommodate diverse blocks and bypass extensive hyperparameter tuning. We applied this method to genotyping data from Project MinE, and achieved 99% average test reconstruction accuracy-i.e. minimal information loss-while compressing the input to nearly 10% of the original size. We demonstrate that haplotype-block based autoencoders outperform linear Principal Component Analysis (PCA) by approximately 3% chromosome-wide accuracy of reconstructed variants. To the extent of our knowledge, our approach is the first to simultaneously leverage haplotype structure and DNNs for dimensionality reduction of genetic data.Availability and implementationData are available for academic use through Project MinE at https://www.projectmine.com/research/data-sharing/, contingent upon terms and requirements specified by the source studies. Code is available at https://github.com/gizem-tas/haploblock-autoencoders.
| S-EPMC11208726 | biostudies-literature