The practice of medicine is fast becoming a data science, and nowhere is this more apparent than with one of Precision Driven Health’s foundation projects – Epidemiology and the estimation of long-term surgical mortality.
Principal investigator Dr Doug Campbell1 from Auckland District Health Board and his team are combining large national datasets of surgical operation and patient demographics with specialist databases such as the national cancer register, to test new modelling techniques for determining surgical mortality.
Dr Campbell believes that the current way of assessing surgical risk is flawed, as the method is based on a metric that only takes into account in-hospital or one-month mortality after surgery. Clinicians have traditionally thought that longer-term mortality is predominantly related to co-existing disease or age, so surgical outcome hasn’t been previously considered as an important factor.
Dr Campbell’s team are challenging this assumption, and they expect their project will reveal that the number of deaths in New Zealand associated with surgery is actually 5500 or more a year, as opposed to the current official figure of around 1800 deaths per year.
The first stage of the project has been to combine the datasets to create one year of national data that will comprise around 300,000 anonymised patients’ data. Using this data, the team will test new modelling techniques for known risk factors such as age, acuity, and grade of surgery. In addition, the American Society of Anaesthesiologists’ physical and cancer statistics have been incorporated into the model, along with novel methods to account for competing and time-dependent risks.
The intention is to develop an easy-to-use surgical outcome calculator that will more accurately describe risk than the tools available to clinicians today. This will enable healthcare providers and patients to be better informed about the risks of surgery, and may result in an overall reduction in expensive – and potentially harmful – surgical procedures. In addition to improving the process of informed consent, the calculator will see better shared decision-making and triaging of high-risk patients and assist in quantifying healthcare performance.
One group that is set to benefit from a more accurate surgical outcome calculator is Māori. Today when Māori undergo surgery, their mortality within one month is higher than for non-Māori. This project would be the first to provide more accurate data about long-term surgical mortality risk, as it will take into account ethnic background along with other factors such as age and gender.
The surgical risk calculator will also have widespread appeal in markets outside New Zealand. While commercial surgical risk calculators are currently sold to district health boards, as well as the National Health Service in the U.K., the tools are flawed because they don’t take into account competing risks and long-term outcomes.
1Dr Campbell is assisted by Dr Tim Short, Associate Professor, Auckland District Health Board; Dr Alan Merry, Professor of Department of Anaesthesiology, University of Auckland; and Luke Boyle, Data Scientist, Orion Health.