Institutional members access full text with Ovid®

Share this article on:

Patient Socioeconomic Status Is an Independent Predictor of Operative Mortality

Bennett, Kyla M. MD*; Scarborough, John E. MD*; Pappas, Theodore N. MD*; Kepler, Thomas B. PhD

doi: 10.1097/SLA.0b013e3181f2ac64
Original Articles

Objective: To evaluate the impact of patient socioeconomic status (SES) on operative mortality within the context of associated factors.

Summary of Background Data: Outcomes disparities among surgical patients are a significant concern. Previous studies have suggested that the correlation between SES and outcomes is attributable to other patient- or hospital-level explanatory factors such as race or hospital wealth. These studies have typically focused on a single explanation for the existence of these inequalities.

Methods: Analyzing more than 1 million records of the Nationwide Inpatient Sample, we used multimodel inference to evaluate the effects of socioeconomic predictors on surgical mortality.

Results: Using univariate and multivariate logistic regression, we find that patient's SES is a strong predictor of operative mortality. Multivariate regressions incorporated many additional hospital- and patient-level covariates. A single-level increase in patient SES results in a mean decrease in operative mortality risk of 7.1%.

Conclusions: SES at the level of the individual patient has a statistically significant effect on operative mortality. Mortality is greatest among patients in the lowest socioeconomic strata. The effect of patient SES on mortality is not mitigated by other explanatory hospital- or patient-level factors.

Patients of low socioeconomic strata experience higher operative mortality rates. This effect is not attributable to other patient- or hospital-level explanatory factors. SUPPLEMENTAL DIGITAL CONTENT IS AVAILABLE IN THE TEXT.

From the Departments of *Surgery and †Biostatistics and Bioinformatics, Duke University Medical Center, Durham, NC.

Reprints: Kyla M. Bennett, MD, Department of Biostatistics and Bioinformatics, Duke University Medical Center, 2424 Erwin Rd, Suite 1102, Durham NC 27710. E-mail:

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (

© 2010 Lippincott Williams & Wilkins, Inc.