Modeling Collective Intelligence in Health Care

Fruitful results after ten years on the road

James Fleck, MD, PhD: Anticancerweb 24(01), 2022

Collective intelligence is a group achievement that results from collaboration and reveals itself as consensual decision-making. It resembles the fractal pattern seen in nature. It is expressed in the form of an emergent property, as a new entity appears as a consequence of the interaction of the parts to create a distinct and expanded whole. As a fractal it is unlimited in growth potential, ordering the components in a natural, beautiful and well-orchestrated iteration. In medicine, collective intelligence is expressed in everyday practice, as the effectiveness of a proposed clinical intervention is the result of a successful interaction between doctor, patient and family. The same happens in medical education, as the relationship between the teaching and student groups enhances the result of pedagogical interventions. The result is fascinating and unpredictable, feeding back growth and producing a high level of personal and social satisfaction. The growth process is so amazing that when participants look back, they don't recognize a pre-established route. When the human mind follows the laws of nature, there is no need for external guidance, using complex mathematical or statistical designs. The methodology has a spontaneous life and always surprises us with its comprehensive efficiency.

When I started this project, ten years ago, I just had the feeling that most of what I have learned in medicine came from the relationship with patients. As a professor of medicine, during this time, I also realized that I was learning from my students. Consequently, collective intelligence became a natural way of life in my dual professional career. Later, I realize the presence of an unexpected new protagonist, the cyberspace. This overgrowing amount of data reflected the infinite connections observed in the human brain. The Internet has become an integral part of people's daily lives, shaping human behavior, habits and skills. Gradually, I improved my skills by approaching companies and highly qualified IT professionals who collaborated in the development and implementation of new digital tools to be used in the practice and teaching of medicine. As I work with cancer patients and teach clinical oncology at my University Hospital, I started this project by creating an educational website that would speak the universal language of cancer epidemiology, diagnosis and treatment.

Very early on, I started to learn from reader feedback. Surprisingly, the apparent semantic complexity of medicine was quite understandable by patients, as long as it was supported by a rational and practical approach. When dealing with medical issues, oversimplification is a deadly mistake, as it distorts the essence of knowledge. This setback was overcome, through precise language and judicious use of explanatory sentences. Progressively, I realized that I was using the same semantics for patients and medical students. Both responded well, but with different motivations. Patients wanted to be cured and students wanted to learn. Nobody was set apart. In fact, we were practicing a broad and open method of inclusive knowledge translation. Later on, several clinical and surgical specialists also appeared among the readers, which further expanded the scope. The text structure was gradually refined. The task was carried out through small editorials, each one requiring a maximum reading of five minutes. The editorial always included a main information, a reasoned explanation, the technical-scientific context in which the data were located and the impact on knowledge translation. Fortunately, the software was responsive, contributing to the majority of readers, who used their cell phones, and the time spent commuting to access the texts. Many of the editorials were about the current molecular signature in cancer diagnosis and treatment, opening a new window to understanding tumor biology and its implications for clinical practice. Patients and clinicians learned about tumor plasticity and better understood the downstream signaling pathway that regulates cell morphology and function. This molecular universe was integrated into clinical simulations, which took the reader to the everyday scenario of health care. Clinical simulations were written to cover the most prevalent malignant tumors. Each clinical simulation was divided into six PLOT elements. In each element of the PLOT, the doctor-patient relationship generated emotions, taking place at different time intervals. Actions were responsible for unique emotional flows, where the deepest inflections represented the patient's turning points. Graphic illustrations were provided. An indexed term browser was developed and used by readers to select a specific tumor or clinical simulation they would like to read, taking a lead role and further personalizing knowledge acquisition. At that moment, a database was created, promoting technical dialogue between students and professionals and supporting patients in sharing decision-making.

Today, the entire project includes two websites, a weekly video production and a massive open online course. The numbers have reached 500 publications, more than 1M accesses and 5K interactions. This information was collected and analyzed resulting in a satisfaction rate of 98.5%. The use of artificial intelligence and deep learning methodology would help in the exploratory search for a standard ontology. Collective intelligence would better categorize and guide the decision-making process in the expansive and limitless database that currently guides cancer diagnosis and treatment. Having achieved this objective, the model could be used in any medical specialty.






3.     Fleck, J: Cancer molecular, clinical and social signature: Massive Open Online Course (MOOC) in Clinical Oncology Version 8.0

5.     You Tube: Grand Round Videos Link:

4.     Photo by Aaron Burden on Unsplash