Patient stratification and identification of adverse event correlations in the space of 1190 drug related adverse events

Roitmann, Eva; Eriksson, Robert; Brunak, Søren

doi:10.3389/fphys.2014.00332

Cited by 21 publications

(15 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2014 described this problem, where the author’s condensed similar trajectories from structured diagnosis codes for the entire Danish population10. However, there are methods that have been found to be useful in exploring adverse drug effects, suicide risk, disease severity and patient stratification in EHR353637383940, these methods depends strongly on the availability of structured information41424344.…”

Section: Discussionmentioning

confidence: 99%

Analysis of free text in electronic health records for identification of cancer patient trajectories

Jensen

Soguero-Ruíz

Mikalsen

et al. 2017

Sci Rep

114

View full text Add to dashboard Cite

With an aging patient population and increasing complexity in patient disease trajectories, physicians are often met with complex patient histories from which clinical decisions must be made. Due to the increasing rate of adverse events and hospitals facing financial penalties for readmission, there has never been a greater need to enforce evidence-led medical decision-making using available health care data. In the present work, we studied a cohort of 7,741 patients, of whom 4,080 were diagnosed with cancer, surgically treated at a University Hospital in the years 2004–2012. We have developed a methodology that allows disease trajectories of the cancer patients to be estimated from free text in electronic health records (EHRs). By using these disease trajectories, we predict 80% of patient events ahead in time. By control of confounders from 8326 quantified events, we identified 557 events that constitute high subsequent risks (risk > 20%), including six events for cancer and seven events for metastasis. We believe that the presented methodology and findings could be used to improve clinical decision support and personalize trajectories, thereby decreasing adverse events and optimizing cancer treatment.

show abstract

Section: Discussionmentioning

confidence: 99%

Analysis of free text in electronic health records for identification of cancer patient trajectories

Jensen

Soguero-Ruíz

Mikalsen

et al. 2017

Sci Rep

114

View full text Add to dashboard Cite

show abstract

“…Previous reports have used clinical notes and narratives, or other electronic health care data, to facilitate the early identification of drug-drug interaction signals or to determine AE co-occurrence [16,17]. For example, in patients with RA, annotation analysis of unstructured clinical notes and subsequent mining of the resultant annotations has been successfully used to calculate the risk for myocardial infarction in those treated with rofecoxib [14].…”

Section: Discussionmentioning

confidence: 99%

“…This approach may involve text mining, which enables efficient analysis of unstructured data elements from clinical narratives [15,19]. Indeed, text mining of clinical sources, among them electronic health care data and clinical notes and narratives, has been used in recent large-scale analyses of drug safety and diseases [16,17,20], including those in patients with RA [21].…”

Section: Introductionmentioning

confidence: 99%

Prodromal signs and symptoms of serious infections with tocilizumab treatment for rheumatoid arthritis: Text mining of the Japanese postmarketing adverse event-reporting database

Atsumi

Ando²,

Matsuda³

et al. 2017

Modern Rheumatology

View full text Add to dashboard Cite

Objective: To search for signs and symptoms before serious infection (SI) occurs in tocilizumab (TCZ)-treated rheumatoid arthritis (RA) patients. Methods: Individual case safety reports, including structured (age, sex, adverse event [AE]) and unstructured (clinical narratives) data, were analyzed by automated text mining from a Japanese post-marketing AE-reporting database (16 April 2008-10 April 2015) assuming the following: treated in Japan; TCZ RA treatment; 1 SI; unable to exclude causality between TCZ and SIs. Results: The database included 7653 RA patients; 1221 reports met four criteria, encompassing 1591 SIs. Frequent SIs were pneumonia (15.9%), cellulitis (9.9%), and sepsis (5.0%). Reports for 782 patients included SI onset date; 60.7% of patients had signs/symptoms 28 days before SI diagnosis, 32.7% had signs/symptoms with date unidentified, 1.7% were asymptomatic, and 4.9% had unknown signs/ symptoms. The most frequent signs/symptoms were for skin (swelling and pain) and respiratory (cough and pyrexia) infections. Among 68 patients who had normal laboratory results for C-reactive protein, body temperature, and white blood cell count, 94.1% had signs or symptoms of infection. Conclusion: This study identified prodromal signs and symptoms of SIs in RA patients receiving TCZ. Data mining clinical narratives from post-marketing AE databases may be beneficial in characterizing SIs. ARTICLE HISTORY

show abstract

“…We calculated pairwise p-values via Fisher's exact test, 4 optionally adjusting for multiple comparisons using the family-wise error rate (FWER) Bonferroni correction or the false discovery rate (FDR) Benjamini-Hochberg correction, both of which have been used in the CNA literature (Roque et al 2011;Roitmann et al 2014;Bhavnani et al 2015;Bagley et al 2016;Kim et al 2018).…”

Section: Link Determinationmentioning

confidence: 99%

Sensitivity and robustness of comorbidity network analysis

Brunson¹,

Agresta²,

Laubenbacher³

2019

Preprint

View full text Add to dashboard Cite

Background: Comorbidity network analysis (CNA) is an increasingly popular approach in systems medicine, in which mathematical graphs encode epidemiological correlations (links) between diseases (nodes) inferred from their occurrence in an underlying patient population. A variety of methods have been used to infer properties of the constituent diseases or underlying populations from the network structure, but few have been validated or reproduced.Objectives: To test the robustness and sensitivity of several common CNA techniques to the source of population health data and the method of link determination. Methods:We obtained six sources of aggregated disease co-occurrence data, coded using varied ontologies, most of which were provided by the authors of CNAs. We constructed families of comorbidity networks from these data sets, in which links were determined using a range of statistical thresholds and measures of association. We calculated degree distributions, single-value statistics, and centrality rankings for these networks and evaluated their sensitivity to the source of data and link determination parameters. From two open-access sources of patient-level data, we constructed comorbidity networks using several multivariate models in addition to comparable pairwise models and evaluated differences between correlation estimates and network structure.Results: Global network statistics vary widely depending on the underlying population. Much of this variation is due to network density, which for our six data sets ranged over three orders of magnitude. The statistical threshold for link determination also had strong effects on global statistics, though at any fixed threshold the same patterns distinguished our six populations. The association measure used to quantify comorbid relations had smaller but discernible effects on global structure. Co-occurrence rates estimated using multivariate models were increasingly negative-shifted as models accounted for more effects. However, only associations between the most prevalent disorders were consistent from model to model. Centrality rankings were likewise similar when based on the same dataset using different constructions; but they were difficult to compare, and very different when comparable, between data sets, especially those using different ontologies. The most central disease codes were particular to the underlying populations and were often broad categories, injuries, or non-specific symptoms.Conclusions: CNAs can improve robustness and comparability by accounting for known limitations. In particular, we urge comorbidity network analysts (a) to include, where permissible, disaggregated disease occurrence data to allow more targeted reproduction and comparison of results; (b) to report differences in results obtained using different association measures, including both one of relative risk and one of correlation; (c) when identifying centrally located disorders, to carefully decide the most suitable ontology for this purpose; and, (d) when relevant to the interpr...

show abstract

Patient stratification and identification of adverse event correlations in the space of 1190 drug related adverse events

Cited by 21 publications

References 41 publications

Analysis of free text in electronic health records for identification of cancer patient trajectories

Analysis of free text in electronic health records for identification of cancer patient trajectories

Prodromal signs and symptoms of serious infections with tocilizumab treatment for rheumatoid arthritis: Text mining of the Japanese postmarketing adverse event-reporting database

Sensitivity and robustness of comorbidity network analysis

Contact Info

Product

Resources

About