95 GOSH legacy and epic inpatient data show changing patient case mix and patient requirements since 2000. (15th December 2021)

• Record Type: Journal Article
• Authors: Bryant, William A; Issitt, Richard W; Briggs, Lydia; Key, Daniel; Booth, John; Shah, Mohsin; Spiridou, Anastassia; Sebire, Neil J
• Is Part Of: Archives of disease in childhood. Volume 106(2021)Supplement 3
• Page Start: A35
• Page End: A35
• Publication Date: 2021-12-15
• Subjects: Children -- Diseases -- Periodicals; Infants -- Diseases -- Periodicals; 618.920005
• DOI: 10.1136/archdischild-2021-gosh.95
• Languages: English
• ISSNs: 0003-9888

Abstract

Abstract : GOSH has a rich patient record spanning many years, but until recently it has not been readily accessible in a coherent dataset. Work by the DRE data engineering team has produced a very simple process to access such datasets with complex cohort building capabilities extremely quickly. With these capabilities comes a unique opportunity to inform current hospital operations for optimisation and development. However, given the size and scope of the data there is a need to understand its structure and how it might be of use in pertinent clinical and operational research. In this work, we have extracted historic GOSH inpatient data across Epic and legacy EPR systems reaching back 21 years to 2000, including ward stay, episode and demographic information. This has allowed us to reconstruct the 'patient resource profile' of the hospital: the distribution of bed resources required for patients by year since 2000. We show that across various specialties the profile of patient bed requirements has notably shifted over the last 20 years. In addition, by combining day case, ICU bed day and non-ICU bed day requirements in a simple resource intensity model, we show that due to the highly skewed distribution of requirements, costs for 'high resource requirement' patients account for a very large fraction of the resource spent in the Trust. This work shows some of the considerations that are required for planning hospital bed use at a high level and illustrates the potential of the very large datasets available through the GOSH DRE team for researchers in the Trust.