Dr. Fernando Suarez R.I.P. In Memory of his Wife and Children

Development and Application of a Prediction Model to Determine the Risk of Atypical Hemolytic Uremic Syndrome and Spinal Muscular Atrophy in Patients Visiting San Ignacio University Hospital.
‍

‍Dr. Fernando Suárez Obando

‍Principal Investigator

‍

San Ignacio University Hospital

Carrera 7° # 40-62, Bogotá, Colombia

1. What are orphan or rare diseases?

In Colombia, an orphan disease is one that is chronically debilitating, severe, life-threatening, and with a prevalence (the measure of all individuals affected by a disease within a particular period) of less than 1 per 5,000 people. This includes rare diseases, ultra-orphan diseases, and neglected diseases. (Law 1392 of 2010/Law 1438 of 2011).

‍

2. What is artificial intelligence?

Artificial intelligence (AI) is a branch of computer science that aims to develop systems capable of "thinking" and "learning" in the same way humans do, to perform complex tasks such as autonomous decision-making, automatic language translation, process automation, and robotics, among others.

‍

3. What are the advantages of using artificial intelligence in medicine?

AI is increasingly used in medicine for its ability to process large amounts of data, identify patterns, and make accurate and rapid diagnoses. It is also used in the research of new therapies and treatments and to improve the efficiency of healthcare systems.

‍

4. What are the advantages of using artificial intelligence in medicine?

It involves finding patients with orphan or rare diseases using natural language processing (NLP). NLP analyzes clinical records in reference centers searching for text patterns that identify patients with diseases of interest, interpreting the written language of doctors to identify patients with a high probability of having a low-prevalence pathology.

5. How is artificial intelligence integrated into medicine concerning orphan diseases?

The low prevalence of orphan diseases makes diagnosis difficult. AI can analyze clinical records for patterns that suggest diagnoses, doing so on a large scale and using the doctors' case descriptions to analyze high volumes of data quickly and accurately.

‍
6. What is the goal of the research study?

To identify patients with orphan diseases by analyzing the text of clinical records.

‍
7. What is the step-by-step process for a patient when entering this study?

After running the AI algorithm on clinical records, those with the highest probability of presenting a case of interest are selected. The patient is then invited for clinical evaluation and confirmatory laboratory tests, validating the AI model through clinical assessment.

‍
8. Why is it so difficult to diagnose an orphan disease?

These diseases have low prevalence, and most doctors are unaware of them. Additionally, confirmation requires very specific biochemical and genetic tests.

‍
9. What happens if a patient has an orphan disease but is not diagnosed in time?

Orphan diseases have management options and, in many cases, specific pharmacological treatments. Without a diagnosis, the patient does not receive appropriate management, causing the disease to worsen.

‍
10. Why is another support method needed for diagnosing orphan diseases?

Due to doctors' lack of awareness, patients experience a diagnostic odyssey, spending years visiting different doctors without a diagnosis, delaying treatment. New strategies are needed to reduce the time to diagnosis.

‍
11. Why use an artificial intelligence model for diagnosing these diseases?

It is a method that allows the analysis of large volumes of clinical data quickly and with high precision.

‍
12. What benefits does early diagnosis of these orphan diseases through artificial intelligence bring to patients and their families?

By identifying patterns that indicate diagnoses, a diagnosis is quickly reached, allowing appropriate management, prognosis, follow-up, and offering suitable genetic counseling for the patient and their family.

‍
13. What benefits does early diagnosis of these orphan diseases through artificial intelligence bring to the healthcare system?

By shortening diagnosis times, complications from untreated diseases are reduced. The healthcare system can effectively address the disease, avoiding unnecessary diagnostic tests and imaging, focusing on required specialists, and eliminating the diagnostic odyssey that generates unnecessary costs.

‍
14. Does artificial intelligence replace the doctor's autonomy?

No, it is a tool that helps better understand clinical patterns and shortens diagnosis times. The doctor's judgment evaluates and validates the AI results.

‍
15. How does this new diagnostic model support the multidisciplinary healthcare team?

The information generated from clinical records and diagnostic images is vast. AI enables quick analysis for any specialty. Additionally, the scientific information on orphan diseases is extensive. Integrating AI as an analysis tool benefits all specialties involved in managing orphan diseases.

‍
16. Has this artificial intelligence model already diagnosed patients with orphan or rare diseases?

Yes, the model has detected patients with rare diseases. These include patients with clinical presentations of the disease or patients already diagnosed, which the model could recognize.

‍
17. Can the artificial intelligence model detect patients already diagnosed with orphan diseases?

Yes, the model can identify previously diagnosed cases.

‍
18. What benefit do patients identified by the model as already diagnosed with the disease receive?

It allows resuming clinical management, proposing new treatments, conducting follow-ups, and updating clinical or imaging studies. Patients with orphan diseases often have poor follow-up; re-identifying them allows appropriate clinical management to resume.

‍
19.Can the diagnostic model help improve the quality of life and mortality of patients with orphan diseases?

In the long term, yes. Identifying the diagnosis and consequently resuming appropriate clinical management impacts the quality of life of the patient and their family.

‍
20. What has been the patients' opinion about knowing and participating in the study?

They have been surprised by the novelty of the approach and have appreciated that their clinical history has been analyzed thanks to the new technology, regaining hope of obtaining a diagnosis.

‍
21. What opinions do patients have about the link between medicine and artificial intelligence?

They are very positive about the possibilities that open up for medicine. Today, patients inform themselves in various ways; the internet opens possibilities, and they are aware that medicine has and uses IT tools that can help improve clinical care.

‍
22. Has there been any impact on awareness of these diseases among the healthcare team due to the study?

Yes, the use of AI incorporates a new approach to orphan diseases. It is a necessary tool given the number of such diseases (about 7,000) and the need to achieve more diagnoses more quickly.

‍
23. Has the study's development impacted medical education?

Yes, medical schools need to understand AI, its advantages and limitations, assimilate the new technology, and build criteria to validate the information derived from the analyses. A medical school must incorporate these tools in education and research.

‍
24. In the future, will this research support medical science for more in-depth genetic analysis?

Yes, genetic tests must be adequately focused on diagnoses. The AI results improve test performance. Additionally, analyzing genetic test results requires AI to evaluate laboratory findings.

‍
25. Could the prediction model be used to diagnose other diseases? Which ones?

Yes, the model is robust and flexible enough to look for patterns of all types of diseases, including those with higher prevalence. For example, cancer, diabetes, myocardial disease, among others. Clinical records "hide" signals of diseases that could be identified and move to predictive identification, helping doctors intervene earlier.

‍
26. What impact do these models have on hospitals and insurers?

Identifying patients quickly and accurately implies timely care, treatment, prognosis, and planning, clarifying the management scenario for the hospital institution and the insurer by recognizing and detailing an epidemiological profile within their population that does not overlook orphan diseases.

‍
27. Who benefits from the prediction models?

Primarily, patients, who will have a more detailed prognostic perspective, implementing preventive management guidelines. Likewise, doctors gain a long-term perspective on the disease, plan management, and accompany the patient throughout the process, optimizing resource use and benefiting the healthcare system.

‍