The Library
Development of risk prediction models combining routine EHR data for use in colorectal cancer screening referral decisions
Tools
Tools
Tools
Cooper, Jennifer Anne, Marshall, Tom, Ryan, Ronan, Parsons, Nicholas R., Stinton, Chris and Taylor-Phillips, Sian (2018) Development of risk prediction models combining routine EHR data for use in colorectal cancer screening referral decisions. In: Methods for Evaluation of medical prediction Models, Tests And Biomarkers (MEMTAB) 2018 Symposium, Utrecht, The Netherlands, 2-3 Jul 2018. Published in: Diagnostic and Prognostic Research, 2 (Supplement 1). pp. 1-44. ISSN 2397-7523. doi:10.1186/s41512-018-0036-3
|
PDF
WRAP-poster-abstracts-2018.pdf - Published Version - Requires a PDF viewer. Available under License Creative Commons Attribution 4.0. Download (1896Kb) | Preview |
|
|
PDF (Poster)
WRAP-development-risk-prediction-Cooper-2018.pdf - Other - Requires a PDF viewer. Download (821Kb) | Preview |
Official URL: https://doi.org/10.1186/s41512-018-0036-3
Abstract
BACKGROUND:
Risk prediction models which incorporate the FOBT with other colorectal cancer risk factors have demonstrated increased sensitivity compared with FOBT alone. EHRs from primary care have a rich level of data and may add a further dimension to risk prediction models. The aim of the study was to determine the availability of GP data for key predictors of colorectal cancer in the screening population and whether we can use this additional information to make more accurate screening referral decisions.
METHODS:
The Health Improvement Network (THIN) database was used to define a screening population by identifying practices which receive electronic bowel cancer screening programme notifications. A prediction model combining the FOBT with other clinical predictors was developed using Cox Regression and multivariable fractional polynomials with backwards elimination. For internal validation, optimism adjusted performance metrics were determined using boot- strapping and absolute risk predictions were estimated.
RESULTS:
The screening cohort derived from THIN gave 292,168 patients. The Cox Regression model which included the FOBT result (n=98,303, 1197 colorectal cancer/polyps) had 13 predictors and 2 interactions including; MCV, various symptoms/diagnoses and whether previous polyps had been diagnosed. The optimism adjusted performance metrics gave a; C-statistic of 0.850, c-slope of 0.991, D statistic 2.298 and R 2 of 0.558. A model investigating negative results only (n = 95,792, 587 colorectal cancer/polyps) included a similar pat- tern of variables. Performance metrics included a C-statistic of 0.650, C-Slope of 0.944, D statistic 0.836 and R 2 of 0.144.
CONCLUSIONS:
This study has shown that a screening cohort can be derived from a primary care database using the electronic bowel cancer screening programme notifications. The prediction models estimate an individual’s absolute risk of colorectal cancer. Additional data could be drawn from primary care onto the Bowel Cancer Screening system using the NHS Spine to contribute to a referral algorithm
| Item Type: | Conference Item (Poster) | ||||||
|---|---|---|---|---|---|---|---|
| Subjects: | R Medicine > RC Internal medicine > RC0254 Neoplasms. Tumors. Oncology (including Cancer) | ||||||
| Divisions: | Faculty of Medicine > Warwick Medical School > Health Sciences Faculty of Medicine > Warwick Medical School |
||||||
| Library of Congress Subject Headings (LCSH): | Colon (Anatomy) -- Cancer -- Prevention, Rectum -- Cancer -- Prevention, Medical screening | ||||||
| Journal or Publication Title: | Diagnostic and Prognostic Research | ||||||
| Publisher: | BioMed Central Ltd. | ||||||
| ISSN: | 2397-7523 | ||||||
| Official Date: | 2 July 2018 | ||||||
| Dates: |
|
||||||
| Volume: | 2 | ||||||
| Number: | Supplement 1 | ||||||
| Page Range: | pp. 1-44 | ||||||
| Article Number: | 12 | ||||||
| DOI: | 10.1186/s41512-018-0036-3 | ||||||
| Status: | Peer Reviewed | ||||||
| Publication Status: | Published | ||||||
| Access rights to Published version: | Open Access | ||||||
| Description: | Poster 12 |
||||||
| RIOXX Funder/Project Grant: |
|
||||||
| Conference Paper Type: | Poster | ||||||
| Title of Event: | Methods for Evaluation of medical prediction Models, Tests And Biomarkers (MEMTAB) 2018 Symposium | ||||||
| Type of Event: | Conference | ||||||
| Location of Event: | Utrecht, The Netherlands | ||||||
| Date(s) of Event: | 2-3 Jul 2018 |
Request changes or add full text files to a record
Repository staff actions (login required)
 |
View Item |
Downloads
Downloads per month over past year
