Resource Management in the Emergency Department by using Machine Learning

The aim of the PhD project is to design a decision support tool based on advanced statistical methods, machine learning/artificial intelligence (AI) and registered health care data in order to optimize emergency department care.

In a present project, an AI based clinical decision support for diagnosing adults with breathing difficulties in the emergency department is being modelled.

The PhD project is part of the research project AIR Lund (Artificially Intelligent use of Registers), a collaboration between Lund University, Halmstad University, Region Halland and Region Skåne. Funded by Region Halland´s Scientific Advisory Board and Sparbanksstiftelsen.

Sweden has an uniqe opportunity for research with register data. With this data, we aim to develop a clinical decision support based on statistical models and machine learning, an aspect of artificial intelligence (AI).

The decision support has potential to improve diagnosis and treatment of patients seeking care at emergency departments. Better and more individualized predictions has the potential to improve patient safety and optimize emergency department resource management.

What is this project and why are you asking me to participate?
Many patients seek emergency care due to breathing difficulties.
The most common conditions causing breathing difficulties in adult patients at an emergency department are acute exacerbation of heart failure, acute exacerbation of chronic obstructive pulmonary disease (COPD), and pneumonia. Diagnosing breathing difficulties is complex, as it requires weighing many different factors. At the same time, rapid and accurate diagnosis and treatment are crucial, since in some cases the symptom may be caused by a serious condition.

In this project, we aim to investigate whether machine learning (artificial intelligence, AI) can be used to train a computer to recognize patterns in emergency patients’ data—such as past illnesses, medication use, examination results, and social factors—in order to predict which of the most common diseases the patient has that are causing the breathing difficulties. The goal is for patients with breathing difficulties to quickly receive an accurate diagnosis and optimal treatment.

In the project, we will collect data from all adult patients (18 years and older) who sought care for breathing difficulties at an emergency department in Region Halland from July 1, 2017 to December 31, 2019, or in Region Skåne from January 1, 2017 to December 31, 2018. Data is collected from Region Halland's and Region Skåne's medical records systems, as well as from the registers of the National Board of Health and Welfare and Statistics Sweden. This includes, among other things: age, sex, socioeconomic data, emergency department visit times, medication treatments, diagnoses and procedure codes, and possible date and cause of death.

All data is protected by confidentiality according to the Swedish Public Access to Information and Secrecy Act (2009:400) and the Secrecy Ordinance (2009:641). The processing of personal data will be conducted in accordance with the General Data Protection Regulation (EU Regulation 2016/679, GDPR). All data processing and analysis will be conducted using pseudonymized data, meaning no individual can be directly identified. All presentation of results will be done at the group level in non-identifiable form, meaning no individual person will be distinguishable. The results will primarily be presented in scientific journals.

Lund University, as the principal institution for the study, is responsible for the personal data and operates under the GDPR. According to the law, you have the right to receive information free of charge about which personal data concerning you is being processed. You also have the right to request corrections to any data you believe is inaccurate.

If you wish to exercise any of your rights, have general questions about how personal data is handled, or would like to request a data extract, you should contact your respective region:

Region Halland
Data Protection Unit, Region Halland, Box 517, 301 80 Halmstad
Phone: 035-13 10 00
Email: [email protected]

Region Skåne
Data Protection Officer, Region Skåne, 291 89 Kristianstad
Phone: 044-309 30 00 (data protection inquiries)
For data extracts: Records and Archive Services, Region Archives, Porfyrvägen 20, 224 78 Lund
Phone: 0771-86 66 00

If you do not want your data to be used, or if you would like more detailed information about the studies, the Secrecy Act, or the GDPR, you are welcome to contact the responsible researcher, Professor Jonas Björk, via PhD student and resident physician Ellen Tolestam Heyman, Emergency Department, Halland Hospital, Halland Hospital Varberg, Box 510, 432 81 Varberg.
Email: [email protected]

If you choose not to allow your personal data to be used, this will in no way affect your current or future contact with healthcare.