"It's essential to combine artificial intelligence with human intelligence, experience and common sense"

Dr Yolima Cossio is a specialist in preventive medicine and public health. In her role as director of Information Systems and Management Support at Vall d'Hebron Hospital, she is currently spearheading the institution's digital transformation. The hospital is already using artificial intelligence software to improve management and patient care, and is working towards becoming a fluid hospital, where the knowledge of professionals and technology can travel without physical barriers to provide excellent care in any type of environment. As for her relationship with the UOC, Dr Cossio is a member of the eHealth Centre's Advisory Board.

As a research centre, what role should the eHealth Center play in addressing global health challenges?

I think that the UOC eHealth Centre has great potential to contribute to addressing global health challenges through innovation and the application of digital technologies. From my experience as a systems director in a hospital, I know that the eHealth Centre's mission to develop and apply new technologies such as artificial intelligence, big data, IoT and telemedicine is crucial to improving clinical aspects such as diagnosis and treatment. But these technologies will also be essential to improve the accessibility, equity and efficiency of the health system and health in general.

I therefore see its role not in developing innovative technologies, but in producing evidence so that these technologies can be implemented in real life. This means demonstrating how they can be effectively integrated into daily practice and health policy. It's not just about testing the accuracy of devices, but showing how they can improve health and make the health system sustainable in real contexts.

How do you think your experience as a physician and data expert in the health sector can add value to the eHealth Center?

The eHealth Center's multidisciplinary approach is essential. The integration of technology with disciplines such as public health, psychology and sociology allows health problems to be approached from multiple angles. This collaboration between researchers, technologists, healthcare professionals and patients is essential to co-create effective and user-centred solutions.

Education and training of health professionals in the use and adoption of digital technologies is another key issue. It's crucial in promoting the uptake of new technologies in their daily practice and to prepare them for emerging challenges.

The fact that the eHealth Center works with governments and international organizations to implement policies that integrate digital technologies effectively and ethically is undoubtedly an added value.

What are the main challenges for a hospital like Vall d'Hebron and how can digital technologies help to overcome them?

One of the main challenges is the ability to meet the growing demand for healthcare services in a context of limited resources, while maintaining the quality of care and meeting the expectations of increasingly empowered users. The high costs associated with chronic diseases, the price of new therapies and the shortage of healthcare professionals are well-known challenges.

Digital technologies can help address these challenges through disease prevention and health promotion programmes, which must be the cornerstone of any healthcare system. In the hospital environment, the digitalization of processes such as virtual consultations, patient monitoring using wearables and the integrated use of artificial intelligence in medical histories offer great potential to improve current challenges. It's important to stress that these technologies must be integrated into a comprehensive strategy and policy of change that goes beyond simple technological implementation.

What is the aim of Vall d'Hebron's digital transformation?

The aims of Vall d'Hebron Hospital's digital transformation are focused on improving people's health and quality of life through the use of technological tools, process improvement and cultural change. These aims include extending the hospital beyond its physical walls, creating a "fluid hospital" that uses digital solutions to go where the knowledge of its professionals is needed. The strategy also aims to create a continuous cycle of learning by using data to improve clinical and management decisions, ultimately transforming practices and processes to be more efficient and effective.

What are the key points of the hospital's big data strategy?

Key points of its data strategy include improving the quality of data at their source, ensuring they are unique and verifiable. Another point involves centralizing the data architecture to ensure data accessibility, interoperability and reuse, the so-called FAIR principles: findable, accessible, interoperable and reusable. We also want to democratize data by making them more accessible to all users, while fully adhering to legal and ethical standards.

Could you give some examples of data projects that have been successfully applied at Vall d'Hebron?

One of the greatest achievements at the Vall d'Hebron hospital complex has been the organization and governance of data, which has allowed various data projects to be successfully implemented. The hospital has focused its efforts on improving the registration, identification and mapping of data sources, as well as data curation and the development of tools that make it easier for users to access unique and credible data for management and decision-making. This approach ensures the credibility and consistency of the information used at operational and strategic levels. A specific example of the successful use of data is the implementation of real-time dashboards for bed management.

How do you ensure the privacy of patient data?

We have developed a data standardization and anonymization tool, which is essential for secondary use of the data. This tool enables researchers to access high-quality data while protecting patient privacy, and to conduct studies with greater rigour and confidence. The foundations have also been laid for specific projects involving advanced algorithms and artificial intelligence techniques.

What are the main roadblocks when it comes to implementing these projects?

Although one might think that the main difficulties in implementing data projects at Vall d'Hebron Hospital are of a technical nature, the reality is that the main challenges have been change management and strategic alignment. The integration of multiple systems and the standardization of data are crucial, but can be overcome with the right technology. However, driving effective digital transformation requires a significant cultural shift within the organization. This includes ongoing training for healthcare professionals and adaptation of organizational processes, which can be met with initial resistance and requires committed leadership.

What are the key benefits of artificial intelligence for a hospital like Vall d'Hebron?

AI can provide significant improvements in the diagnosis and treatment of patients, for example by increasing diagnostic accuracy, particularly in areas such as radiology and pathology, allowing diseases to be detected at an earlier stage. Similarly, AI can personalize patient care by recommending treatment plans that are tailored to each individual's unique characteristics, using data based on their medical history, genetics and other relevant factors. AI can also optimize hospital resource management and decision-making based on predictive outcomes.

And the potential risks?

What I fear most about the implementation of AI is the possibility of bias and inequity that this technology can create. If AI algorithms are trained on biased data, they can perpetuate and amplify these inequities in diagnostic and treatment decisions.

Another significant risk is technology dependency, which leaves us vulnerable in the event of connectivity or power outages, and the hospital unable to operate if over-reliant on these systems.

How can we overcome the bias that large amounts of data can sometimes contain?

Overcoming bias in large amounts of data requires a combination of artificial intelligence, intelligence, experience and common sense. Diversifying data sources is critical to ensure they represent a wide range of populations and situations. Data cleaning and pre-processing play a critical role, as they can identify and mitigate biases before the data are used to train algorithms. In addition, implementing machine learning techniques that are transparent and explainable allows developers to identify and correct biases at an early stage. Ongoing monitoring and evaluation of models under development ensures that biases are not perpetuated or exacerbated. Finally, fostering a diverse team of developers and data experts provides different perspectives and helps identify and correct biases that might otherwise go unnoticed.

How are professionals adapting to new technological changes?

In my experience, adaptation to the new healthcare reality behaves like Rogers's innovation adoption curve. There are professionals who are quick to adopt technologies and integrate them into their daily practice, while others are naturally sceptical. The vast majority, however, fall somewhere in the middle. They are eager to see the results and willing to jump on the bandwagon if there is good leadership.

Finally, how are patients responding to the hospital's digital transformation?

The response of patients to the digital transformation of the hospital is one of the issues that concerns us most. During the pandemic, we dispelled many prejudices by seeing how quickly our patients adapted to digital solutions out of necessity. However, some patients find it difficult to accept this change in mindset, especially if they see no immediate benefit. Our strategy therefore focuses on involving patients in the design and implementation of digital solutions, while respecting their right to face-to-face care if they prefer.