Evaluating Quantile Regression for Health Care--Associated Infections Gap Time Data

• Activity Number: 124
• Type: Contributed
• Date/Time: Monday, August 4, 2014 : 8:30 AM to 10:20 AM
• Sponsor: ENAR
• Author(s): Jonathan Edwards*+
• Companies: CDC
• Keywords: Healthcare ; Infection ; Quantile Regression

Abstract

CDC's National Healthcare Safety Network (NHSN), used by CDC and its partners for surveillance of healthcare-associated infections (HAIs), provides hospital performance measures on HAI incidence to help promote healthcare quality. Supplementary measures are needed to differentiate hospitals with recurring HAIs. Beginning in 2011, continuous reporting of central line-associated bloodstream infections (CLABSIs) commenced in NHSN allowing for analysis of recurring CLABSI gap times within hospital ICUs. Quantile regression models based on hospital and unit-level factors were estimated.

In 2011, there were 77,587 months of data summarizing CLABSI incidence reported to NHSN from 6,775 hospital ICUs. Hospital-level CLABSI summary incidence measures are heavily influenced by exposure volume and event rarity in CLABSI recurrence. Quantile regression analysis was used to estimate conditional quantiles and survival functions that can be used to augment existing measures and provide additional help to distinguish CLABSI recurrence among hospital ICUs and provide support to performance quality measurement.