Impact of Multidrug Resistant Organisms on Length of Stay in Patients with Catheter-Associated Urinary Tract Infections and Central Line Associated Blood Stream Infections

Background:

Healthcare associated infections (HAIs) result in increased length of hospitalization, excess morbidity and healthcare costs.  The relative contribution of multidrug resistant infections (MDROs) on such variables is less well documented.  In 2008, mandatory HAI reporting using the National Healthcare Safety Network (NHSN) was initiated in all 255 Pennsylvania acute care facilities.  Although all HAIs are reported in PA, a smaller number are used to calculate infection rates within and between facilities (requiring accurate denominator information).  These include catheter-associated urinary tract infections (CAUTIs) and central line-associated bloodstream infections (CLABSIs).  We used the first 9 months of reported data to assess the influence of MDRO on length of stay (LOS) for CAUTI and CLABSI in PA. 

Objective:

To compare the impact of infection with an MDRO versus a non-MDRO on the LOS for persons with CAUTI and CLABSI in PA hospitals. 

Methods: Data for the period July 2008 to March 2009 on CAUTI and CLABSI were examined.  For ease of analysis, ward types were collapsed into two strata:  critical care units (CCU) (excluding neonatal intensive care and special care) and general wards.  Infection rates, device utilization ratios (DUR) and length of stay (LOS) were calculated for both CAUTI and CLABSI.  Length of stay was examined for admission to event, event to discharge, and admission to discharge.  Data were stratified for the two ward types based on presence or absence of an MDRO. Student t-tests that assume unequal variances were used to compare MDRO-related to MDRO-unrelated CAUTI or CLABSI for DUR and LOS.

Results: DURs were higher in critical care patients for both CAUTI (0.64) and CLABSI (0.55) than in patients admitted to the general wards (0.21 and 0.20), but were similar for patients with and without an MDRO infection for both CAUTI and CLABSI.  The period of admission to event was longer than that of event to discharge for patients with both CAUTI and CLABSI. The presence of an MDRO resulted in significantly longer length of stay for patients with CAUTI in both the CCU (28.6 days vs 21.6 days) and general ward (21.3 days vs. 15.4 days), but did not appear to have a similar impact for patients with CLABSI. 

Conclusions: Although this represents <1 year of data, length of stay appears to be influenced by whether or not infection is due to MDRO in CAUTI but not in CLABSI, and may be driven by the timing of event onset rather than response to treatment.  Further analysis is necessary to assess these findings and to evaluate the role of prior antimicrobial usage.