Dataset Information

Classification of clinically useful sentences in clinical evidence resources.

ABSTRACT: Most patient care questions raised by clinicians can be answered by online clinical knowledge resources. However, important barriers still challenge the use of these resources at the point of care.To design and assess a method for extracting clinically useful sentences from synthesized online clinical resources that represent the most clinically useful information for directly answering clinicians' information needs.We developed a Kernel-based Bayesian Network classification model based on different domain-specific feature types extracted from sentences in a gold standard composed of 18 UpToDate documents. These features included UMLS concepts and their semantic groups, semantic predications extracted by SemRep, patient population identified by a pattern-based natural language processing (NLP) algorithm, and cue words extracted by a feature selection technique. Algorithm performance was measured in terms of precision, recall, and F-measure.The feature-rich approach yielded an F-measure of 74% versus 37% for a feature co-occurrence method (p<0.001). Excluding predication, population, semantic concept or text-based features reduced the F-measure to 62%, 66%, 58% and 69% respectively (p<0.01). The classifier applied to Medline sentences reached an F-measure of 73%, which is equivalent to the performance of the classifier on UpToDate sentences (p=0.62).The feature-rich approach significantly outperformed general baseline methods. This approach significantly outperformed classifiers based on a single type of feature. Different types of semantic features provided a unique contribution to overall classification performance. The classifier's model and features used for UpToDate generalized well to Medline abstracts.

SUBMITTER: Morid MA

PROVIDER: S-EPMC4836984 | biostudies-literature | 2016 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Classification of clinically useful sentences in clinical evidence resources.

Morid Mohammad Amin MA Fiszman Marcelo M Raja Kalpana K Jonnalagadda Siddhartha R SR Del Fiol Guilherme G

Journal of biomedical informatics 20160113

<h4>Unlabelled</h4>Most patient care questions raised by clinicians can be answered by online clinical knowledge resources. However, important barriers still challenge the use of these resources at the point of care.<h4>Objective</h4>To design and assess a method for extracting clinically useful sentences from synthesized online clinical resources that represent the most clinically useful information for directly answering clinicians' information needs.<h4>Materials and methods</h4>We developed ...[more]

PMID: 26774763

Similar Datasets

Project description:Comparative effectiveness research (CER) calls for substantial changes in the way clinical research is conducted, interpreted and practically applied in the USA, in order to produce useful clinical evidence. Departing from classic efficacy and effectiveness research, the evolving CER paradigm requires structural and substantive innovations that address three basic questions: what works? for whom? and in whose hands? Addressing these questions will require fundamental changes in the approach to clinical research that include: the use of active treatments (or comparators) versus placebos in the comparisons of treatments, innovative or 'alternative' research methods, the specification and a priori design of studies to account for important subgroups, accounting for the nested nature of healthcare delivery in design and analysis of CER, the simultaneous study of multiple treatments or treatment modalities, the study of multiple outcomes (benefits and harms) for each treatment compared, and the reassessment of the value of different study designs in the hierarchy of collective 'evidence'. In order to aid individual providers and patients in making informed, personalized treatment decisions, guided by the best evidence possible, CER studies must generalize to a broad range of subgroups reflecting the spectrum of patients, providers and health systems that populate real-world practice settings. Without expansion in the scope, conduct and subsequent interpretation of clinical research reflected in the issues outlined above, CER will fall short of its potential for informing evidence-based practice and personalized medicine. The current paradigm for conducting, interpreting and applying clinical research does not meet the needs of optimal generalizability and application to individual physician-patient efforts to identify the most effective treatment, and therefore does not support the basic requirements of CER. The proposed changes should neither require decades nor exorbitant budgets to achieve. Using two examples, prostate cancer, and comparisons of single medications, we illustrated how the proposed changes in clinical research, matching strategy to each application, might be addressed.

Dataset Information

Classification of clinically useful sentences in clinical evidence resources.

Publications

Classification of clinically useful sentences in clinical evidence resources.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets