The pandemic has shown us this: guaranteeing access to treatment during emergencies is crucial, and extremely complex. During peaks in access to health facilities, it is essential to be able to better manage beds and available staff, to avoid overcrowding of the wards and the overload of work for doctors and nurses, which result in waiting, frustration and even potential health damage. for patients. The problem, which arose well before Covid 19, cannot be solved by simply increasing the resources available: it is not just a question of how many beds are available in a ward but also of how they are used. Distributing the flows of patients, both within the single hospital and the entire territorial network, foreseeing and anticipating their needs, is the key to avoiding over-requests and to guaranteeing targeted care in the shortest possible time. In this direction, as Philips, a leading company in the Health Technology sector explains, help comes from predictive models based on artificial intelligence algorithms: digital assistants who can study health data, present and past, collected by a hospital or a network of hospitals, and provide forecasts that support health professionals in a more informed and aware decision-making process. By allowing, for example, to predict which and how many patients in the emergency room will have a good chance of being admitted to intensive care. Philips talks about what the patient journey could be like with the new coordinated approach.
Anticipating the moves
Rethinking the management of patient flows through information tools based on predictive analysis is the key to improving the accessibility and quality of care, to the benefit of both the patient and the healthcare staff. Let’s imagine that Mrs Rosa suddenly has palpitations and breathing difficulties and calls the ambulance quickly. Let’s assume that, at this point, an operator who is responsible for coordinating patient flows receives a notification with information on the patient’s health status. The operator will thus be able, based on the needs of the person and on the data relating to the availability of the various structures, to direct Rosa to where there is room. This is to avoid overcrowding in emergency rooms as well as a delay in diagnosis and treatment, which can have serious consequences. In the event of an overload of requests, for example due to an exceptional event, the coordinator quickly obtains an overview and will be ready to promptly contact experts who act as supervisors within a larger network to activate emergency plans and increase places and staff. Once Rosa arrives at the hospital, clinical information, combined with data obtained through artificial intelligence algorithms, allows for predictions on the risk of deterioration. In this way the operator can pre-allocate a place for the patient within the right ward, also considering the resources and medical tools necessary for the treatment.
An overview
Predictive models based on artificial intelligence algorithms therefore make it possible to avoid congestion in the departments, and to treat a greater number of people and in a more targeted manner. Having an overview, thanks to centralized operational information, improves the management of patient flow and the organization of health care within the entire company. Furthermore, predictive analyzes allow to speed up and optimize the path of taking charge and care of the patient, which saves travel and time spent in hospital. That freed time and space will thus be available to other people who need treatment.
Decision-making is extended to the whole network
It is not enough to manage everything through a command center, or a centralized structure, used to control cases. It is necessary to coordinate and make information available to all staff in the area. In particular, this is possible through the choice and definition of specific KPIs (Key Performance Indicators), essential indicators of the progress of a company in achieving the set objectives. These parameters must allow a reliable prediction of “bottlenecks”, critical issues that in the event of an increase in the flow of patients can generate significant slowdowns, as well as provide synthetic information, useful for the decision-making process of the most appropriate interventions. Once the KPIs have been defined, these data must be made available at the single assistance point – therefore disseminated and dislocated – and be usable in a simple way by all staff members. In this process, the role of the doctor remains central: predictive algorithms can favor more coordinated choices within the entire care network, but clinical decisions, which take these new elements into account, are and will always be the responsibility of the specialist.
Integrated management even at home
Coordination and integrated patient management does not end in the hospital. Monitoring, also based on predictive analyzes, also continues remotely once the person returns home. This constant control serves to recognize any alarm bells, for example a possible worsening of some parameters, to act promptly and prevent more serious complications. The hospital also benefits from this: emergencies are avoided and hospitalization costs are reduced. One pilot study conducted in the United States, for example, has shown that with this method it is possible to have an 80% reduction in the second hospitalizations of patients with chronic obstructive pulmonary disease (Bpco). This approach can also be useful in the long term, through more informed and informed planning of the resources and type of care supported by health facilities.