For many hospitals around the world, surgical demand is outstripping capacity as the number of patients visiting hospital continues to climb at an alarming rate. This is leading to patients experiencing long delays in surgery as doctors and surgeons struggle to meet the demand.

This overwhelming demand makes planning and prioritising surgery a huge challenge for hospital management teams. With limited resources, surgical targets must be realistic to be achieved. It is incredibly difficult for hospitals to realistically plan and project throughput when demand is so high.

Counties Manukau Health (CM Health) is a leading District Health Board (DHB) in New Zealand and is eager to improve services. CM Health covers a region in the North Island of New Zealand making up 11% of the country’s population, a large, highly diverse population.  They have contributed to several improvements to the national health system in collaboration with the other DHBs and are determined to become the best healthcare system in Australasia.

However, CM Health’s two surgical centres are struggling to meet their surgical targets, while also having difficulties with accurate estimation of how close they will get to their targets. The planning for CM Health is based on socio-demographic information, projected population and information about how health services are used – such as specialty surgeries. After engaging with the Precision Driven Health (PDH), a project was established to help CM Health plan and prioritise surgeries in a more efficient and data driven manner. This project aims to build accurate predictive models of surgical throughput. The driving purpose behind the project is to help the executive management team at CM Health balance their targets, so they can decide how many surgeries of certain specialties can be achieved each day.

Luke Boyle, data scientist at Orion Health, worked closely with the ORUA research group at the University of Auckland, particularly PhD student Tom Adams and his supervisors senior lecturer Dr Michael O’Sullivan and associate professor Cameron Walker, to build an optimisation model to help surgeons and managers meet targets. This model simulates surgical throughput to gain an understanding of how the surgical centres would be able to achieve their targets. The model also suggests optimal session allocations. This helps managers decide how to assign specialties to sessions, while getting as close to targets as possible.  The model was built on two years of historical surgical data allowing accurate prediction of what sort of patients and operations CM Health are likely to see. The tool uses this data to more accurately predict how often each type of patient and surgery would present to CM Health, as well as how long each operation would take.

The model provides an overview of how to allocate sessions, rather than specifying the type of surgery at each hour of each day. The model was built this way as some resources, such as specialty surgeons, are not always available all day, every day and a model that is too specific would be difficult to use in a meaningful way.

The surgical planning and prioritisation tool developed by Precision Driven Health can be used as evidence of what can be achieved, as well as providing suggestions about the optimal way to allocate limited resources. This brings a higher level of efficiency and accuracy to surgical planning. The surgical planning and prioritisation tool is now able to be used as an evidence-based, data driven resource to help management teams plan for the next financial year.

This project reflects the meaningful ways data science can be used to help healthcare organisations optimise their resources and budgets, delivering the most efficient care to their patients.