ABSTRACT
Background: Real-world evidence through secondary use of data (SUD) in oncology is gaining increasing interest, to better understand cancer epidemiology and provide insights into treatment patterns in daily practice. This study evaluates incidence of HR+/HER2- early BC (eBC) and its management in clinical practice through SUD and gauge the impact of the SARS-COV2 pandemic. Method(s): This observational retrospective analysis integrates administrative databases for healthcare resources consumption (pharmaceuticals, hospitalizations, diagnostic tests and specialist visits databases) from a sample of Italian Local Health Units, based on 15 million inhabitants across Italy. Patients with >=1 hospitalization discharge diagnosis for BC, with surgical intervention and HR+ status (determined by coding for HR+ status or by presence of endocrine therapy) between 01/2010-06/2021 were included. Patients with at least one prescription of anti- HER2 monoclonal antibodies were excluded. Patients were classified by menopausal state through prescription for the gonadotropin-releasing hormone analogues (GnRHa). Incidence was calculated during all study period. Result(s): Incidence rate has a slight upwards trend, as expected, ranging from 53.9 in 2013 to 62.7 in 2019 per 100,000 health-assisted subjects. Incidence in 2020 is 49.2 per 100,000 (table 1 for quarter split). As for adjuvant therapies, 31,836 patients were included in the analysis of which 5343 (16.8%) were classified as premenopausal. Mean age was 64.5 years. Most patients (78.8%) were treated with only adjuvant endocrine therapy (ET). 16.5% of the sample received adjuvant chemotherapy (CT). CT treatment was more prescribed in premenopausal patients. CT treatment was started within 12 weeks of surgery for 3.9% of the sample. Most patients (12.7%) started it between 12 weeks and 24 weeks. Conclusion(s): SUD can provide lots of information with the right queries. The analysis confirms the slight increase in incidence observed by national registries and provides an estimate of the impact of SARS-COV2 with a 22% reduction of breast surgery in 2020. Administrative data can be used to assess clinical variables (e.g. premenopause through GnRHa prescription), and could be further explored for disease stage through axillary dissection, and recurrence through prescription of therapies used in metastatic setting.
ABSTRACT
Objectives: To estimate the prognostic factors underlying severity of Sars-Cov-2 infection using a machine learning approach. Methods: The analysis is based on administrative databases of Italian Entities. Patients who were hospitalized with COVID-19 diagnosis (ICD-9 078.89) after 1st January 2020 were included into the dataset together with 13 relevant features representing age, sex and clinical history of each patient. Each record was labelled as 0 (hospitalized patients) or 1 (patients in intensive care or deceased). KerasTuner was used to define the architecture of the Neural Network achieving good accuracy score. To identify prognostic factors underlying severity of Sars Cov-2 infection, feature’s importance was evaluated starting from a Random Forest Classifier. Results: The preliminary dataset built contains 10.448 records from 9.346 hospitalized patients. The selected neural network is made of 13 input nodes, each one representing a feature, 1024 nodes in the hidden layer, processing information that comes from the input layer, and 2 nodes in the output layer, each one representing a label to define patient’s condition. The neural network obtained was able to achieve 64% of accuracy on the testing set. The condition of approximately 2 out of 3 patients was correctly predicted just by analysing their features. The feature’s importance computed from the Random Forest Classifier indicated that patient’s age is the primary prognostic factor underlying severity of Sars Cov-2 infection. The combination of the other features slightly improved model’s performance. Conclusions: The preliminary analysis shows that age is a prognostic factor of fundamental importance in defining the severity of Sars Cov-2 infection. The model obtained could be used to predict disease progression in patients most at risk by analysing their information in the databases. The model will be further improved through a process of feature selection to increase its accuracy and to allow the identification of other prognostic factors.