Developed in-house, the Hstory platform aggregates critical patient data to provide analytical reports to doctors within minutes

18/06/2024

---

Sidney Klajner — Foto: Silvia Zamboni/Valor

A few days ago, a patient visiting surgeon Sidney Klajner was taken aback during his consultation. Before the patient could say anything about his medical history, the doctor began detailing past medical issues, diagnoses, and surgeries. “He asked me how I knew so much without him saying a word,” said Mr. Klajner, president of the Albert Einstein Israelite Hospital and a specialist in digestive system and coloproctology. The answer? Generative artificial intelligence.

While the use of AI in healthcare is not new, it has traditionally been focused on management, utilizing algorithms to enhance internal processes and improve institutional efficiency. Now, artificial intelligence is making its way into the consultation room, transforming the interaction between doctors and patients.

Einstein Hospital is launching an AI platform called Hstory, internally conceived and developed. It scans medical records, extracts the most relevant data, and presents it in minutes as an analytical report. The data can be organized chronologically, including all events regardless of medical specialty, or by body organ.

With Hstory, Einstein is the first hospital in Brazil to launch a platform that combines AI and Big Data for direct patient care, and one of the few worldwide. Stanford University School of Medicine was the first to launch such a system in January.

“There was a need to engage healthcare professionals, whose training is generally very traditional,” said Mr. Klajner. Given this background, many doctors view technology with skepticism. By demonstrating that AI can also offer benefits in patient care, the expectation is to overcome healthcare professionals’ resistance and pave the way for innovation. “From an adversary, the doctor becomes an ally,” said Mr. Klajner.

Time-saving is the primary advantage of Hstory. To make a reliable diagnosis, doctors must conduct lengthy interviews with patients and consult various systems for test results and other critical data. By automating this task, doctors can reallocate the time saved to other aspects of the consultation. Mr. Klajner reports having conducted a simulation using his own medical records and saving 20 minutes with Hstory.

The new platform is the result of a long-term investment plan. Efforts began in 2016 when Einstein’s directors traveled to the United States to observe other hospitals’ experiences and conceive their own project. This led to the creation of a Big Data department, which has grown significantly over time. From an initial team of six, it now comprises around one hundred members.

Asking the right questions is crucial in choosing the path forward, said Mr. Klajner. One question that Einstein’s doctors pondered was why many patients were readmitted shortly after discharge.

The hospital created an algorithm to determine whether patients were adhering to prescribed diets and medications at home, among other variables, which helped identify weak points and improve discharge procedures, said neuroradiologist Edson Amaro Junior, head of Einstein’s Big Data department.

The algorithm increased the likelihood of identifying patients at high risk of readmission within 30 days by four times and led to enhanced preventive actions for these patients. It also freed up space for 34 more short-duration surgeries per month by optimizing the use of beds and operating rooms. Another effect was a 20% reduction in “office time”—spent on bureaucratic tasks—for nursing teams, freeing up 83 hours of work per month, according to the hospital.

Einstein has already developed over one hundred Big Data algorithms, with 20 for prediction. Currently, 49 systems use AI across 16 areas of the institution. Of these, 18 systems are in use, while the rest are in the testing phase.

Hstory is now available to all doctors with access to Einstein’s electronic medical record system. This includes about 3,000 professionals, including those working at the public Vila Santa Catarina Municipal Hospital, managed by Einstein in partnership with the Unified Health System (SUS).

Despite rapid advances in AI in medicine, technology will not replace the doctor’s role in critical decisions such as treatment choices, medication prescriptions, or surgery recommendations, said Mr. Klajner. “This practice will not be replaced. Artificial intelligence will not do any of that.”

The future of AI in healthcare poses strategic challenges that companies, governments, and regulatory agencies must address. One such challenge is the interoperability of digital systems across institutions, which would allow for the creation of a single electronic medical record. When a patient arrives at a hospital, even for the first time, the doctor would have complete access to their history, including data from all hospitals and clinics they have visited.

Achieving this level of integration depends on public and private hospitals investing in Big Data and digital security and establishing common technological standards. Another consideration is who would finance the infrastructure necessary for this connectivity. “Israel has had a unified medical record for years, but it has 9.5 million inhabitants, not 200 million like Brazil,” said Mr. Klajner. “The future must always be pursued.”

*Por João Luiz Rosa — São Paulo

Source: Valor International

https://valorinternational.globo.com/