ICU infectionsSunday, April 01, 2012, 14:54 - 15:06
Prevalence and economic impact of hospital-acquired infections in intensive care units: retrospective analysis from a USA hospital databaseF. Joly*, A. Stemart, C. Amand-Bourdon, E. Obi-Tabot (Chilly-Mazarin, FR; Bridgewater, US)
Objectives: To determine the prevalence of three types of hospital-acquired infections (HAI) known to be the drivers of high economic impact in intensive care unit (ICU) from a U.S. hospital database.
Methods: A retrospective cohort study was undertaken using hospital database from the Premier Perspective of adults (>=18 yrs old) admitted to an ICU with a stay >= 48 hrs in 2007, 2008, and 2009. The three HAI assessed were: bloodstream infection (BSI) including sepsis, surgical site infection (SSI), and hospital-acquired pneumonia (HAP) of ventilator-associated pneumonia (VAP). The economic impact was assessed using inpatient mortality rate, length of stay (LOS) and the cost for inpatients (encompassing all actual costs to treat a patient including all supplies, labour, and depreciation of equipment).
Results: Data were collected on 463, 491 patients with 511,815 stays (Mean, 1.10 stay/patient) in ICU following the entry criteria in 2007 of whom 91.5 % had only one stay over the year. Patients in ICU were mostly elderly (53.3%) who were often concerned of device use as mechanical ventilation (21.9%) and central catheter (26.2%), which are high risk factors for HAI. Among the stays, 26.7% were concerned of at least one HAI with the following prevalence: 16.9% for HAP/VAP, 14.5% for BSI, 39.7% for sepsis and 1.5% for SSI. The inpatient-mortality rates were higher for HAP/VAP with 16.7%, followed by 13.9% for BSI and 10.9% for SSI. SSI showed the highest LOS (23.4 days + 23.4), followed by HAP/VAP (15.2 days + 16.6) and BSI (12.7 days + 13.9). One-day ICU costs (USD) were similar throughout the three HAI (BSI: 2,621 + 1857, SSI: 2,582 + 1296 and HAP/VAP: 2,362 + 1198). Globally, an increase in inpatient mortality was observed in patients with HAI (13.2% vs. 1.1%) and an increase in LOS (12.5 days vs. 6.9 days) compared to patients without HAI. However, costs/day were less in stays for patients with HAI (2583.9 USD), compared to stays for patients without any HAI (2965.9 USD). The same incremental differences were noted for each specific infection and the same trend was demonstrated for 2008 and 2009.
Conclusions: The prevalence and burden of HAP/VAP, BSI and SSI in ICU are high and are clearly associated with a higher economic impact. The increase in mortality rates and longer LOS findings as the drivers of higher ICU costs, indicate the need for specific measures to reduce the prevalence of these major types of hospital-acquired infections.