The hospital as a data-driven organisation


Like any other organisations in this day and age, hospitals and other health care providers now have access to increasingly large amounts of data. They can potentially augment electronic patient records with data from patients’ own devices: wearables, mobile phones and other gadgets. Far more hospital equipment now contains sensors, monitoring everything from the patient to the condition of the machine itself. Machine-to-machine connectivity can pass information to clinicians rapidly and effectively, and alert them to even the smallest change in a patient’s condition.

This avalanche of data offers huge potential to provide useful information to support decision making and improve patient outcomes. From which patients to prioritise at which point, right through to support for diagnosis and treatment, clinicians are likely to see advanced algorithms, such as those driven by artificial intelligence (AI), supporting their day-to-day decisions. In time, it seems likely that AI-driven algorithms will become as ubiquitous as X-rays and MRIs in hospitals.

A willingness to embrace technology

Health care providers and clinicians have never been slow to use technology to improve patient outcomes. They have, naturally, sometimes held back because of cost implications – MRI scanners are not cheap, for example. But they have always been quick to see the potential of new technology and to use it to improve patient care.

AI, however, has been slower to take off. Somehow, many hospitals and health care providers do not seem to be ready for decision-making supported by algorithms. Perhaps it is a change of culture and a concern about the explainability of decisions supported by a "black box." Perhaps staff simply do not yet have the necessary skills and experience to take advantage of the insights locked into data. Whatever the reason, it has been a fairly slow start.

There is now, however, a groundswell towards data-driven decision making. A number of health care organisations have started to embrace AI and analytics. They have often begun with small-scale projects, but there is growing recognition that the future lies in personalised health care – and that this depends on data and analytics.

AI offers a unique combination of quality and safety for patients, better outcomes, and reduced costs. After all, getting the right medication or treatment quicker, with fewer side effects, is significantly cheaper than trying a number of expensive options first. It is also far better for patients.

Looking towards the future

Over the next three to five years, it seems likely that more and more health care providers will start to become data-driven organisations. This will, in most cases, need a change in culture. Providers must move towards acceptance of the process of using data to generate insights that then drive decisions. It is likely that this acceptance will grow as organisations see the early impact.

Providers will need to support the change in culture with changes in three other areas. The first is staff competence in using analytics and understanding the insights that emerge. It is vital that staff understand the recommendations from the decision engine and are able to explain these to patients and other staff. The second area is infrastructure. Hospitals will need suitable facilities and equipment to gather data and then analyse it.

Developing a data strategy

A successful data-driven hospital needs to centralise its data strategy for business operations and care.

This means that health care providers must develop strong data and model governance arrangements. Staff and managers alike need to be sure that data quality is high and the outputs from models remain appropriate. Models are only as good as the data that is fed into them. And insights are only as good as the models.

A successful data-driven hospital needs to centralise its data strategy for business operations and care.

It is not reasonable to expect IT staff to be responsible for data that is input by clinicians. Clinicians, therefore, need to understand the benefits of high-quality data, and take responsibility for ensuring that patient data is correct. This is a bit of a vicious/virtuous cycle. Until people see the benefits of decisions driven by reliable data, it is hard to persuade them that reliable data is important. However, without reliable data, it is impossible to generate the necessary impact. A strong data strategy – covering collection, assurance, preparation and use – will go a long way to help.

There are no shortcuts in this process. It takes time and energy. Cultural change is never easy. In this case, however, it will be helped by the obvious benefits to patients, and a historical willingness in health care to accept new technology that improves patient outcomes.


About Author

Joost Huiskens

Leave A Reply

Back to Top