Risk factors for methicillin-resistant Staphylococcus aureus (MRSA) infection among hospitalized patients with nasal MRSA colonization

Background:  Nasal methicillin-resistant Staphylococcus aureus (MRSA) colonization has been associated with the development of MRSA infection.  Prior studies have elucidated risk factors for the development of nasal MRSA colonization as well as the development of MRSA infection.

Objective:  To identify factors associated with development of MRSA infection among hospitalized patients with nasal MRSA colonization.   

Methods:  We performed a nested case-control study based on a retrospective review of medical records after obtaining approval from our institutional review board.  The study setting was Parkland Memorial Hospital, a 770-bed public, academic, tertiary care hospital in Dallas, Texas.  The study duration was from 1/1/2008-7/28/2009.  Within the cohort of patients with nasal MRSA colonization upon active surveillance, patients who developed MRSA infection after 48 hours but within 3 months of admission [cases] were compared with uninfected nearest admission controls [2 per case].  Patients with MRSA infection on admission or 3 months prior were excluded.  Demographic, clinical, and microbiologic data were collected.  Statistical analyses were performed with SPSS 17®.  All tests were two-tailed and statistical significance was defined as p<0.05.  

Results:  During the 19 month study period, 426 patients were found to be colonized with MRSA.  Of these, 36 (8.5%) developed a subsequent infection with MRSA.  The infections included 17 (47.2%) respiratory infections, 8 (22.2%) soft tissue infections, 5 (13.9%) urinary tract infections, 2 (5.6%) intra-abdominal infections, 2 (5.6%) bacteremias, and 2 (5.6%) other infections.  These 36 cases were compared to 72 matched controls.  Factors associated with development of MRSA infection upon bivariate analysis were a history of steroid use at admission (19.4% vs. 2.8%, p=0.006), development of renal failure (44.4% vs. 25%, p=0.049), breakdown of skin integrity (52.8% vs. 15.3%, p=0.00), antibiotic use after admission (100% vs. 63.9%, p=0.00), mean patient days (2.00 vs. 1.24, p=0.00), ICU days (8.86 vs. 2.17, p=0.00), ventilator days (0.11 vs. 0.06, p=0.00), central line days (3.06 vs. 2.35, p=0.00), number of surgeries (0.92 vs. 0.38, p=0.005), and number of procedures (11.56 vs. 4.72, p=0.00).  When these variables were fitted into a multivariable model, we found that pre-colonization use of steroids (p=0.023) and mean number of patient days (p=0.000) were predictive of MRSA infection.  MRSA infected patients are more likely to be discharged to skilled nursing facility/rehabilitation centers (38.9% vs. 6.9%, p=0.000) compared to uninfected controls.

Conclusions:  Hospitalized patients with nasal MRSA colonization who develop subsequent infection with MRSA are more likely to have received steroids in the 3 month period prior to colonization and are more likely to have a longer hospital stay.