Social Deprivation and Postoperative Mortality

Socioeconomic deprivation is a multidimensional determinant of health that may influence perioperative outcomes through several pathways, including differences in baseline health status, burden of comorbidities, access to timely healthcare, perioperative optimization, and hospital-level resources. Although previous studies have suggested an association between social deprivation and worse postoperative outcomes, most available evidence comes from high-income countries or from studies focused on specific surgical populations. Data from low- and middle-income countries, including Brazil, remain limited.

This study is a retrospective, multicenter cohort study designed to assess the association between socioeconomic deprivation and perioperative outcomes among surgical patients treated in Brazilian hospitals. Participating centers will include both public and private hospitals from different regions of the country. The study will use routinely collected institutional data from patients who underwent surgical procedures during the 2024 calendar year. No additional contact, intervention, or change in clinical care will occur as part of the study.

The main exposure of interest is socioeconomic deprivation, assessed through georeferenced area-level indicators based on each patient's residential address. Residential addresses will be geocoded to identify the corresponding census tract or human development unit. Socioeconomic deprivation will then be measured using validated Brazilian indices, including the Brazilian Deprivation Index, the Social Vulnerability Index, and the Municipal Human Development Index. These indices capture different dimensions of socioeconomic context, including income, education, household conditions, urban infrastructure, human capital, and vulnerability related to work and income.

The study database will include demographic, clinical, surgical, and structural variables routinely available in hospital electronic records. These variables include age, sex, ethnicity, American Society of Anesthesiologists physical status classification, educational level, occupation, surgical procedure, surgical urgency, date of surgery, date of hospital discharge or in-hospital death, surgical specialty, surgical magnitude, and type of healthcare system. Surgical procedures will be standardized using Brazilian procedural classification systems (Classificação Brasileira Hierarquizada de Procedimentos Médicos -CBHPM). If a patient undergoes more than one eligible surgical procedure during the study period, only the first procedure will be considered for outcome analysis.

Data will be extracted locally by the information technology teams of participating centers using structured database queries. Identifiable patient information will not be included in the analytical dataset. Residential addresses will be used only for geocoding and linkage to area-level socioeconomic indices. After geospatial processing, the final research dataset will exclude full residential addresses and other direct identifiers. Access to study data will be restricted to authorized members of the research team, and data will be stored in a secure institutional environment.

The primary analysis will evaluate whether higher levels of socioeconomic deprivation are associated with 30-day in-hospital postoperative mortality. Deprivation indices will be analyzed in categories, including quintiles, to assess gradients of risk across socioeconomic strata. Baseline characteristics will be summarized across deprivation groups using appropriate descriptive statistics. Multivariable logistic regression models will be used to estimate the association between socioeconomic deprivation and postoperative mortality after adjustment for relevant demographic, clinical, surgical, and hospital-level factors. Planned covariates include age, sex, ethnicity, ASA physical status, surgical urgency, surgical magnitude, surgical specialty, and healthcare system type.

Model performance will be assessed using calibration and discrimination measures. Calibration will be evaluated using the Hosmer-Lemeshow goodness-of-fit test, and discrimination will be assessed using the area under the receiver operating characteristic curve. Results will be reported as odds ratios with 95% confidence intervals. A significance level of 5% will be used for statistical analyses.

The study also aims to explore whether incorporating socioeconomic deprivation into perioperative risk assessment may improve the identification of vulnerable surgical patients. By evaluating social deprivation alongside established clinical and surgical risk factors, this study may contribute to a better understanding of inequities in perioperative care and support future strategies for risk stratification, resource allocation, and quality improvement in both public and private healthcare settings in Brazil.