109 COCATOO: a cohort comparison tool for GOSH clinical researchers. (30th November 2020)

• Record Type: Journal Article
• Authors: Briggs, Lydia; Bryant, William; Shah, Mohsin; Booth, John; Issitt, Richard; Spiridou, Anastassia; Sebire, Neil J
• Is Part Of: Archives of disease in childhood. Volume 105(2020)Supplement 2
• Page Start: A37
• Page End: A37
• Publication Date: 2020-11-30
• Subjects: Children -- Diseases -- Periodicals; Infants -- Diseases -- Periodicals; 618.920005
• DOI: 10.1136/archdischild-2020-gosh.109
• Languages: English
• ISSNs: 0003-9888

Abstract

Abstract : When faced with large amounts of data, it can be a challenging and lengthy process for a researcher to identify key differences and trends between different sets of patients. The development of a reporting tool which can be easily adapted to any speciality and patient selection would be advantageous in highlighting and presenting patient group comparisons in a clinical research environment. By developing standard, reproducible analytics from the core datasets provided by the Digital Research Environment (DRE) team at GOSH, the researcher is able to identify differences in hospital (hospital admissions, ward stays) and clinical (patient laboratory results, demographics and vital signs) trends between user defined cohort-control patient groups which can accelerate their clinical understanding. Herein, we introduce COCATOO, a cohort comparison tool that compares and reports distinct sets of patient groups based on defined specifications which are controlled by the researcher. By utilising generalisable analytics developed in R, the tool is able to display a set of predefined exploratory data analytics (EDA) which have the ability to be adapted to any hospital department by incorporating project specific rules. We present here an example with Chronic Kidney Disease (CKD) comparing stage 5 to stages 1–4. The output report gives an overview of the disease and highlights significant cohort-control differences via the predefined EDA.