Occupational diseases are one of the greatest concerns for any worker, as they significantly affect their quality of life and represent an economic burden on companies and governments. This article proposes an artificial intelligence model capable of predicting a worker’s future medical referral, one to two years from now, based on the progress of their medical examinations. To do so, recurrent neural networks with Long Short-Term Memory are used to determine the severity of the referral. The results, although not ideal, indicate a good starting point for the development of more models that allow for personalized prevention of occupational diseases.