Evolving Role and Future Directions of Natural Language Processing in Gastroenterology

Nehme, Fredy; Feldman, Keith

doi:10.1007/s10620-020-06156-y

Cited by 21 publications

(12 citation statements)

References 90 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our present work provides a starting point for building a lexicon of terms and phrases to support Natural Language Processing approaches. 38 PRO measures are increasingly seen as a practical alternative to reliance on composite activity measures generated by clinicians. This is particularly relevant in the face of the current COVID-19 pandemic which has disrupted traditional face-to-face care and enforced the move to remote consultations, with reduced access and/ or willingness to undergo objective testing.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Heterogeneity in outcome assessment for inflammatory bowel disease in routine clinical practice: a mixed-methods study in a sample of English hospitals

et al. 2021

View full text Add to dashboard Cite

ObjectivesKnowledge of the extent of variation in outcome assessment for inflammatory bowel disease (IBD) in routine practice is limited. We aimed to describe and quantify variation in outcome coverage and to explore patient, clinician and practitioner factors associated with it.DesignProspective exploratory mixed-methods study.SettingIBD clinics at six hospitals in North West England with differing electronic health record (EHR) systems.MethodsMixed-methods study comprising: (a) structured observations of outcomes elicited during consultations (102 patients consulting 24 clinicians); (b) retrospective analysis of outcomes recorded in the EHR (909 consultations; 127 clinicians) and (c) semistructured interviews with the 24 observed clinicians. We determined whether specific outcome ‘sets’ were elicited or recorded, including: (1) a minimum set of symptom pairs (‘PRO-2’); (2) symptom sets from disease activity indices and (3) a reference list of 37 symptoms, signs and impacts. Factors associated with variation were explored in univariate and multivariate binary logistic regression analyses and from clinician interviews.ResultsPRO-2 coverage was not invariable (elicited during 81% of observed consultations; recorded in 56% of EHR) and infrequent for complete activity indices (all domains from Harvey-Bradshaw Index: elicited, 18%; recorded, 5%). The median number of outcomes from the reference list elicited per consultation was 12 (13-fold variation) and recorded in EHR was 7 (>20-fold variation). Symptom quantification (PRO-2) seldom adhered closely to standardised descriptors and an explicit timeframe was defined rarely. PRO-2 recording in EHR was associated with a diagnosis of ulcerative colitis (OR: 2.09 (95% CI 1.15 to 3.80)) and nurse-led consultations (OR: 6.98 (95% CI 3.28 to 14.83)) and a three-way model suggested 26% of total variability lay between clinicians, 17% between patients but the remainder was unexplained. Most clinicians expressed preference for individualised health status evaluations versus standardised outcome assessments.ConclusionsThere was little evidence for standardised assessment and recording of IBD outcomes and substantial intra-clinician and inter-clinician variation from one consultation to another. Nurses demonstrated a greater tendency to standardised practice.

show abstract

Section: Discussionmentioning

confidence: 99%

“…Our present work provides a starting point for building a lexicon of terms and phrases to support Natural Language Processing approaches. 38 …”

Section: Discussionmentioning

confidence: 99%

Heterogeneity in outcome assessment for inflammatory bowel disease in routine clinical practice: a mixed-methods study in a sample of English hospitals

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Natural language processing (NLP) has been proposed as a potential tool to retrieve clinically meaningful information from free-text documents in gastroenterology. [10][11][12] The study aimed to evaluate the impact of introducing a direct-access IBD-physician delivered endoscopy pathway for new primary care referrals with suspected IBD.…”

Section: Endoscopymentioning

confidence: 99%

“…Natural language processing (NLP) has been proposed as a potential tool to retrieve clinically meaningful information from free-text documents in gastroenterology 10–12…”

Section: Introductionmentioning

confidence: 99%

Impact of direct-access IBD physician delivered endoscopy on clinical outcomes: a pre-implementation and post-implementation study

Stammers

Rahmany

Downey

et al. 2022

Frontline Gastroenterol

View full text Add to dashboard Cite

IntroductionPatients with suspected inflammatory bowel disease (IBD) referred from primary care often face diagnostic and treatment delays. This study aimed to compare a novel direct-access IBD endoscopy pathway with the traditional care model.MethodSingle centre real-world study analysing primary care referrals with suspected IBD. Group A: patients triaged to direct-access IBD endoscopy. Group B: patients undergoing traditional outpatient appointments before the availability of direct-access IBD endoscopy. Demographics, fecal calprotectin (FCP), C-reactive protein (CRP), disease activity score, endoscopy findings, treatment and follow-up were collected and statistically analysed. Ranked semantic analysis of IBD symptoms contained within referral letters was performed.ResultsReferral letters did not differ significantly in Groups A and B. Demographic data, FCP and CRP values were similar. Referral to treatment time (RTT) at the time of IBD endoscopy was reduced from 177 days (Group B) to 24 days (Group A) (p<0.0001). Diagnostic yield of IBD was 35.6% (Group B) versus 62.0% (Group A) (p=0.0003). 89.2% of patients underwent colonoscopy in Group B versus 46.4% in Group A. DNA rates were similar in both groups. The direct to IBD endoscopy pathway saved 100% of initial IBD consultant clinics with a 2.5-fold increase in IBD nurse-led follow-up.ConclusionOur novel pathway resulted in an 86% reduction in RTT with associated increased diagnostic yield while saving 100% of initial IBD consultant outpatient appointments. Replication in other trusts may improve patient experience and accelerate time to diagnosis/treatment while optimising the use of healthcare resources.

show abstract

“…For the nonstructured text‐format data, machine learning models, after being trained, can now also be deployed automatically and efficiently to label and categorize and even change them to a structured format. Natural language processing technology can extract meaningful information from unstructured data sources, transform it into a structured format in a real‐time fashion, and absolve the labor‐intensive chart review processes 2 . This is of utmost importance, as a huge amount of archived medical data in most healthcare organizations, which is extremely valuable, is actually stored in a nonstructured format, and utilization of such information can enrich the abundance of healthcare big data and extend its application.…”

Section: Clinical Data Format and Application Of Artificial Intelligementioning

confidence: 99%

Application of artificial intelligence in gastroenterology: Potential role in clinical practice

Kuo

Chiu

2021

J of Gastro and Hepatol

View full text Add to dashboard Cite

Artificial intelligence (AI) based on deep learning boosted medical research in the past years and is expected to enormously change the style of health care in many aspects in the foreseeable future. Nowadays, there are exploding volumes of healthcare‐related data being generated daily. Because of its time‐sensitive characteristics, being able to process large amounts of data in real‐time fashion is crucial in healthcare settings. In gastroenterology practice, AI can manage and interpret the sheer amount of data with different formats coming from a myriad of sources, including currently used endoscopic or imaging devices, digital record systems, and electronic health records, or from other sources such as governmental databases, social media, or wearable devices over a long period. Traditional gastroenterology is nowadays beginning to transform to a new personalized, predictive, and preventive paradigm. Evidence‐based practices and recent studies are coming out every day, and big data‐based approaches and the progress in basic sciences and its emerging applications are now becoming the indispensable part of precision medicine. In gastroenterology, AI can be applied in disease diagnosis, treatment guidance, outcome prediction, and reducing workload of the healthcare staff. As the healthcare community begins to embrace AI technology, how to seamlessly construct an interoperable platform to accommodate data with high variety and veracity with high velocity and implement AI in the clinical workflow would be the future challenges.

show abstract

Evolving Role and Future Directions of Natural Language Processing in Gastroenterology

Cited by 21 publications

References 90 publications

Heterogeneity in outcome assessment for inflammatory bowel disease in routine clinical practice: a mixed-methods study in a sample of English hospitals

Heterogeneity in outcome assessment for inflammatory bowel disease in routine clinical practice: a mixed-methods study in a sample of English hospitals

Impact of direct-access IBD physician delivered endoscopy on clinical outcomes: a pre-implementation and post-implementation study

Application of artificial intelligence in gastroenterology: Potential role in clinical practice

Contact Info

Product

Resources

About