How applying analytics and algorithms to health data can generate clinical, operational and financial insights
- By David Labajo Izquierdo, VP - Healthcare Digital, Europe
This time last year, Coronavirus was not even on the health agenda. Its surprise arrival and swift spread throughout 2020 put global health systems to the test. Now we have ridden the first wave of the global pandemic, analysed first responses and are readying for a new crescendo. What was hoped to be a 3-4 month health emergency has now morphed into a long-term focus on recovery and renewal.
On one side, healthcare organisations have a big challenge in catching up on delayed activity such as screening, diagnosis and surgery, to recover from the backlog of missed appointments, and to avoid another potential crisis. But on the other hand, and at the same time, they need to prepare and plan for the future, and a long-lasting set of upcoming COVID-19 waves. There is a growing acceptance that COVID-19 will not be overcome quickly, if ever, and we need to adapt to the new normal and return our focus onto achieving positive patient care. The big question, of course, is how do we do the same, if not more, with the same amount of staff and resources? How to work differently, to not struggle with the new waves? How can we be ahead of the situation, instead of running behind the train?
The answer most probably is about data. Healthcare organisations that will succeed will be the ones able to mix the best clinical practice, with the best data practice and the best human Intelligence, with the best Artificial Intelligence (AI).
Healthcare professionals spend a lot of their time collecting and reporting information from patients, data from patients, that in most of the cases they are not benefiting from nor are they ever using it again. We need to make the data work for the professionals, instead of the professionals working for the data.
The application of analytics or algorithms to health data can generate clinical, operational and financial insights that can certainly be key for a more efficient, effective and sustainable healthcare. The benefits and outcomes are enormous, and we can see them at three levels:
1. At the point of care – embedding AI and Deep Learning into medical equipment to automate and speed up examinations or reporting.
For example, AI inside mammography or other oncology-based review and reporting systems can help prioritise a second review of images by a radiologist that have been identified as having early changes or small lesions. This automated ‘flagging up’ of case images has the potential to save many hours each day and increase the success of an early diagnosis that in many conditions like cancer, are also bringing automatically a better survival rate and a lower treatment cost.
2. At a departmental level - to monitor demand and capacity, optimise equipment or room utilisation and planning staffing levels.
For example, department scheduling can give clear visibility of ICU bed, operation theatre or CT scanner availability to efficiently manage valuable resources. It can also help to address backlogs and understand the factors behind patients missing appointments.
The application of AI to automate non-critical activities, gives back the time for those critical activities where the healthcare professionals are key. AI also gives the capacity to predict in advance the workload in departments. This helps, to size resources and prioritise activity accordingly, resulting in a better use of resources and a reduction in waiting lists, thereby avoiding collapse situations and ending into a more efficient department.
3. From a hospital level – predict and manage flow through the entire hospital ecosystem.
Having an overview of the entire hospital environment is the holy grail of health informatics. Examples already exist of hospital consortiums using AI for capacity planning of lifesaving resources such as ventilators and negative pressure beds. The automatic analysis of millions of data points generates real-time operational insights that can help save lives by matching patients to the resources they need most.
Improving patient outcomes and increasing access to care
Treating patients at the point of care is the first and vital touch-point in the patient’s access to care provision and can be the critical stage in their overall health outcome. Imagine a patient has arrived at A&E: The priority here is to triage the patient swiftly and accurately; diagnostics from lab results, imaging and past medical history will help quickly determine a diagnosis and plan the next stage in their care journey. Clinicians need to accurately decide: does the patient need to be admitted, or can they be discharged to home with community-based follow up?
If requiring admittance, insight is then needed on where there is bed capacity in the appropriate department and a possible theatre or follow-up radiology slot. A dashboard of departmental or entire-hospital capacity helps to plan an efficient journey of personalised care for the patient and deliver productivity within the entire hospital environment.
What the arrival of COVID-19 showed us was that unpredictable events that demand unexpected capacity on a large scale has the potential to overwhelm health systems. Positive patient outcomes are entirely dependent on early diagnosis, swift intervention and thorough care planning. An added value is that the earlier the diagnosis, the lower the total cost of treatment and hospital stay.
Planning and prioritising healthcare workload
Gaining health insights using AI is not just to smooth the patient journey and reduce waiting times. It can help to accurately plan and deploy more effectively the capital equipment and human resources in the health environment. People will always be at the heart of healthcare and having enough at the right time and in the right place is crucial.
Identifying the peaks of activity needing staffing in emergency departments or intensive care units helps to pragmatically plan the deployment of workforces in advance, reducing panic and potentially lowering locum costs.
Creating a calm and stable health environment
Recovering from backlogs and planning for healthcare in the new COVID-era demands a strategic move towards AI. Futureproofing healthcare to deliver a calm and stable environment requires deep information on how to prioritise and plan the time of your people, equipment and resources, together with new ways of monitoring, controlling and diagnosing patients in a variable and changing environment.
Organisations with the best data capabilities and practices i.e., putting together all the value of the patients’ data with AI applications and integrated with the clinical practice, will be the only ones able to manage the healthcare activity in a planned and predictable way. This will deliver the best health, clinical and financial outcomes.