The retrospective study will be used to develop an artificial intelligence model of risk stratification of physiological and psychological complications arising from the information available in the electronic medical record and first consultation report to support patients and healthcare professionals in better managing the healthcare process for patients diagnosed with long COVID.

The stratification of the risk of complications related to persistent COVID both physiological and psychological in a personalized way would optimize the cost-effectiveness model for the management of these patients. Similarly, the early detection of complications associated with persistent COVID in patients belonging to vulnerable groups would improve care times and, therefore, the patient's prognosis.

The primary objective for this study is to gather anonymized retrospective data of patients suffering from long COVID in order to contribute to the generation of the SENSING-AI cohort.

Inclusion Criteria:

• Legal adult
• Diagnosed of long COVID-19 in the last year
• With the presence of any of these symptoms:
• Asthenia (Tiredness)
• Dyspnea
• Shortness of breath
• Anxiety
• Stress
• Depression
• Sleep disorder

Exclusion Criteria:

• Attended to specialized care consultation
• Was admitted in hospital in the last year due to a problem not related to the COVID complications

The sample size for the retrospective study enough to generate a first version of the risk stratification models will be around 100 cases. The target population will be as balanced as possible between subjects who needed specialized care due to long COVID-19 complications (either specialized care consultations or any non-planned hospital admission) at 1 month, 3 months, 6 months and 1 year from the long COVID-19 diagnose and those who did not require such specialized care.