DATA structuring and performance indicators for decision making in oncology and hematology

Abstract Background: Digital transformation in Brazilian hospitals has heightened the demand for high-quality data and strategic indicators to support evidence-based decision-making. However, many institutions still struggle with fragmented data sources, a lack of standardization, and weak data governance, factors that hinder analytical maturity and limit organizational visibility and performance management. In this context, robust data structuring and the development of key performance indicators (KPIs) are essential to generate value and enhance the competitiveness of healthcare services. Considering this, the study aims to evaluate the impact of data structuring and KPI development on operational efficiency and decision-making in a hospital setting. Methods: This retrospective, applied study adopted a mixed-methods approach, integrating quantitative and qualitative analyses. The study was conducted at a philanthropic Jewish hospital in São Paulo, Brazil, specializing in oncology and hematology, from April 2024 to July 2025. Data from institutional databases, a data lake, and spreadsheets stored on a shared drive were consolidated. An automated Extract, Transform, Load (ETL) workflow was implemented using Python scripts to standardize processes and enable daily updates. Relevant data fields were extracted via SQL queries and integrated into Power BI for real-time dashboard visualization. Dashboards were deployed on institutional servers with scheduled refreshes and controlled employee access to support continuous monitoring. Results: The analytics team successfully developed and implemented 55 operational dashboards, improving data infrastructure and KPI tracking to support more effective monitoring and decision-making. Key outcomes included optimizing the occupancy rates of adult and pediatric outpatient infusion units. Upon identifying that the adult unit was operating at full capacity while the pediatric unit had significantly lower occupancy, five pediatric infusion stations were reallocated to adult care. This adjustment balanced resource utilization and increased the number of patients treated daily from an average of 95 to 115, a historical record. In the chemotherapy outpatient unit, initial no-show rates ranged from 10% - 13%. Following the implementation of dashboards and targeted interventions, including inpatient status tracking, proactive appointment confirmations, and cancellation pattern analysis, no-show rates decreased to 3% - 4%, representing a reduction of over 50% and enhancing patient throughput. The implementation of integrated management dashboards enabled oncology and hematology leaders to perform period comparisons, variance analyses, and annual forecasts, providing actionable historical and predictive insights. Additionally, the development of itemized revenue dashboards enabled daily revenue fluctuation tracking through segmentation of grouped billing items, improving financial monitoring and performance analysis. Conclusion: The study demonstrates that effective data structuring and the definition and use of performance indicators are powerful enablers of hospital management. Their adoption fosters faster, evidence-based decision-making and contributes to enhanced operational efficiency and governance. However, achieving these benefits requires investments in technology, staff training, and the cultivation of a data-driven organizational culture.