The wheat from the chaff #50 Rare diseases and their challenges

Hello friends 👋 , 

Rare diseases affect only a small fraction of the population - less than 1 in 2,000 people in Europe or less than 200,000 in the United States. According to Anvisa, these conditions are defined by the absence of a cure and limited treatment options. In Brazil, the National Policy for Comprehensive Care for People with Rare Diseases classifies them as those affecting up to 65 out of every 100,000 individuals. 

Although the criteria vary between countries, they all share a common challenge: difficult diagnosis and restricted treatment. You've probably heard of some of these diseases, such as Pompe disease, cystic fibrosis, Huntington's disease and amyotrophic lateral sclerosis (ALS). 

The diagnosis of these conditions is often a long journey, marked by years of uncertainty and suffering for the patient and their family. Fortunately, advances in technology, such as artificial intelligence and the analysis of large databases (such as DATASUS, in Brazil), have proved to be valuable allies on this path. 

Let's separate the wheat from the chaff 

Chic is simple 

Imagine living with an unknown disease, spending years searching for an accurate diagnosis, having your symptoms confused with other conditions. Until finally a health professional identifies the cause: Pompe disease, a rare neurodegenerative condition. However, the delay in diagnosis has already had a significant impact on this person's quality of life. 

Fortunately, DATASUS, Brazil's public health database, offers a valuable solution in this context. With its comprehensive records on the situation and dynamics of public health in the country, DATASUS has become a key player in helping to locate and follow up patients with rare comorbidities. 

Recently, a major pharmaceutical company set itself a significant challenge: to help locate patients with any of the rare diseases mentioned above. This project involved the use of various data sets, including public information from DATASUS, the company's own internal data and extensive use of technologies such as modeling and artificial intelligence. 

After initial discussions and data analysis, we decided to change our initial strategy. Instead of focusing solely on locating patients, we opted to identify the doctors and health institutions most likely to treat cases of these rare diseases. This approach proved to be more effective, as by properly training the teams at these locations, they would be better prepared to recognize and correctly diagnose these conditions. 

Using data from DATASUS, we analyzed an 11-year period, taking into account ICD (International Statistical Classification of Diseases) codes. By tracking these codes over time, we can identify which ICDs appear most frequently in patients who eventually receive a diagnosis of the rare disease in question. This allows us to better understand the path taken by these patients until they reach the correct diagnosis, what we call the “diagnostic saga”. 

For example, if we take a patient diagnosed with a rare disease and analyze the history of the previous 11 years, our aim is to discover the most common ICD codes during that period and begin to understand the path of the disease. We realize that, during this diagnostic process, the patient ends up being treated for various other conditions, such as arthritis and chronic pain, until they reach the correct diagnosis, in this case, Mucopolysaccharidosis (MPS). 

In a simplified way, we were able to use data and artificial intelligence models to define the most likely previous illnesses until the individual was diagnosed with a rare disease such as Pompe disease. This allows us to follow the patient's progress month by month. 

By developing a predictive model, we identify the most important variables (image below), which allows us to calculate a “score” for each SUS patient (without identifying them individually). In this way, we can act preventively, before the situation worsens. 

This integrated approach, powered by AI, promises to revolutionize healthcare, making it more accurate, efficient and patient-centered. By tracking the entire patient’s journey, everyone involved can make more informed and effective decisions, resulting in better clinical outcomes and more optimal use of available resources. 

The age of AI is just beginning! 

We would love your help sharing this website!

If you know someone that might like our posts, don’t be shy and share it with them! By the way, we appreciate this!

See you Sunday…

We always arrive in your inbox around 12:08 PM. Some email servers are stubborn and delayed... Others are even worse and throw us into spam and/or promotions. What's the logic behind this timing? ...None :)