Dados do Resumo

Título

Can the visualization of integrated data on postoperative outcomes of head and neck surgeries in a Power BI dashboard promote protocol improvements?

Introdução

The increasing volume of data generated in the healthcare sector is transforming how clinical decisions are made through the creation of new data-driven technological solutions. In the field of oncology, data visualization approaches can also be employed. In the context of oncological surgery, this strategy can improve management and clinical outcomes by providing a detailed analysis of postoperative outcomes, such as complications and recovery rates, which can lead to adjustments in surgical protocols and better care outcomes.

Objetivo

To create a dashboard in Microsoft Power BI with data on the outcomes of head and neck surgeries performed at a cancer center in São Paulo city.

Métodos

The research data is part of a pre-existing database from a head and neck surgery department of a Brazilian cancer center, hosted on REDCap. These data correspond to all surgical procedures performed between 2014 and 2023, collected by attending physicians and residents. The data collection process followed a rigorous procedure, with information curated by the department's medical team, primarily through presentations of results in departmental meetings. The database hosted on REDCap was integrated with Microsoft Power BI via an API (Application Programming Interface), allowing one-click import and update of the utilized information, minimizing the chance of errors or data breaches. The research project was approved by the Ethics Committee, opinion number: 6.885.340.

Resultados

The database used contains 19,506 surgical procedures, with 11,744 (60.2%) performed on female patients and 7,762 (39.8%) on male patients. A total of 17,132 (87.8%) procedures were privately funded, while 2,351 (12.1%) were funded by the Unified Health System (SUS). Among the patients, 3,044 (15.6%) were smokers and 2,682 (13.8%) consumed alcohol. During the sample period, there were 365 (1.9%) cases of readmission within 30 days, 630 (3.23%) cases of surgical site infection (SSI) within 30 days, 248 (1.3%) reoperations within 7 days, and 171 (0.9%) postoperative deaths within 7 days.
The final version of the dashboard features a navigation interface that allows users to access four distinct tabs: "patient profile," "general indicators," "thyroidectomies," and "other procedures." In all tabs, data can be filtered and updated instantly to view information for specific dates within the sample period or for each surgeon in the department individually. The "patient profile" tab displays data related to comorbidities, smoking, alcohol consumption, body mass index (BMI), age, among others. The other tabs focus on more specific variables for thyroidectomies, such as hypocalcemia and vocal cord paralysis. Another relevant piece of data included in the dashboard is the number of preoperative decolonization protocols applied to patients undergoing major surgeries.

Conclusões

The results foster important reflections on improving clinical outcomes in oncology. The tool enables the rapid identification of issues and facilitates projects that optimize care protocols. The results can also be used to accelerate the discussion on value-based reimbursement, providing evidence for negotiations with health plans, and encouraging data-driven practices to enhance care quality and improve clinical outcomes.

Palavras Chave

Data Administration; Surgical Oncology; Head and Neck Cancer