¹Governo do Estado do Espírito Santo, Vitoria/ES, Brazil
²ICEPi - Instituto Capixaba de Ensino, Pesquisa e Inovação em Saúde, Vitoria/ES, Brazil

Text

Introduction: Currently, data science is standing out in all data analysis markets, where large amounts of data and information are manipulated. In healthcare, where we need to analyze, predict, diagnose, and treat large populations, it serves as a method to help understand overall health.

Methods: Data was analyzed using data science (SAS Viya – an analysis, artificial intelligence, and data management platform) from a database between 2011 and 2024, involving over 50,000 active public servants (of both genders, over 18 years old, and all types of occupations such as health, education, security, administration, etc.) to establish patterns of absenteeism (work absence due to occupational illness). After the data analysis, smart dashboards were created using Power BI, categorized by type of agency (health, education, security, etc.), position, function, type (ICD), cost (in Brazilian reais), duration (days), and city. Once the dashboard was assembled, it was presented to the managers of the relevant agencies along with a joint action plan to reduce absenteeism rates.

Results: The data analyzed from public employees between 2011 and 2024 showed that the state incurred costs exceeding 1 billion reais due to absenteeism from occupational illness, with mental disorders (anxiety, depression, stress) being the leading cause of absences. The education sector had the highest absenteeism rates for females, while the security sector had the highest for males, with an average absence duration of 9 days.

Discussion: The data analyzed spanned from 2011 to 2024 because prior to this period, the government did not have control over recording the type of absence (ICD) in occupational illness records; this was only suggested starting from that date. Another point of discussion is the gender with the highest absenteeism; while men represented the highest costs, it can be argued that they are a larger proportion of professionals in the state. The total difference in absenteeism costs between men and women was not more than 100,000 reais, suggesting that women may be more significantly impacted by work absences.

Conclusion: Data science helps healthcare manage large amounts of data and information, enabling better decision-making in the evaluation, prevention, diagnosis, or treatment of diseases and health conditions.