Objective The aim of this study was to develop an accurate regional forecast algorithm to predict the number of hospitalized patients and to assess the benefit of the Electronic Health Records (EHR) information to perform those predictions. Materials and Methods Aggregated data from SARS-CoV-2 and weather public database and data warehouse of the Bordeaux hospital were extracted from May 16, 2020, to January 17, 2022. The outcomes were the number of hospitalized patients in the Bordeaux Hospital at 7 and 14 days. We compared the performance of different data sources, feature engineering, and machine learning models. Results During the period of 88 weeks, 2561 hospitalizations due to COVID-19 were recorded at the Bordeaux Hospital. The model achieving the best performance was an elastic-net penalized linear regression using all available data with a median relative error at 7 and 14 days of 0.136 [0.063; 0.223] and 0.198 [0.105; 0.302] hospitalizations, respectively. Electronic health records (EHRs) from the hospital data warehouse improved median relative error at 7 and 14 days by 10.9% and 19.8%, respectively. Graphical evaluation showed remaining forecast error was mainly due to delay in slope shift detection. Discussion Forecast models showed overall good performance both at 7 and 14 days which was improved by the addition of the data from Bordeaux Hospital data warehouse. Conclusions The development of hospital data warehouses might help to get more specific and faster information than traditional surveillance systems, which in turn will help to improve epidemic forecasting at a larger and finer scale.

Aggregated data from 2020-05-16 to 2022-01-17 regarding Bordeaux Hospital EHR. Bordeaux hospital data warehouse was used, during the pandemic, to describe the current state of the epidemic at the hospital level on a daily basis. Those data were then used in the forecast model including: hospitalizations, hospital and ICU admission and discharge, ambulance service notes and emergency unit notes. Concepts related to COVID-19 were extracted from notes by dictionary-based approaches (e.g. cough, dyspnoea, covid-19). Dictionaries were manually created based on manual chart review to identify terms used by practitioners. Then, the number and proportion of ambulance service calls or hospitalization in emergency units mentioning concepts related to covid-19 were extracted. Due to different data acquisition mechanisms, there was a delay between the occurrence of events and the data acquisition. It was of 1 day for EHR data, 5 days for department hospitalizations and RT-PCR, 4 days for weather, 2 days for variants and 4 days for vaccination. For the training and evaluation of the model, the chosen date was the date of data availability to mimic a real-time streaming forecast.

Data are stored in a .rdata file. Please use R (https://www.r-project.org/) software to open the data.