Evaluation of the Effectiveness of Hospital Infection Control Programs (HICP) Using a Benchmark Indicator at 5 Intensive Care Units (ICUs) in Buenos Aires City

Background: Although a valid benchmark represents an important tool for assessing the performance of HICP, these standards may vary over time by changes in methodology applied for surveillance as well as improvements associated with the infection control process. In addition, the representativeness of standards needs to be considered when an individual institution is compared. The development of a summary indicator that could integrate the evolution of device-associated infections (DAIs) rates compared with a selected benchmark over time could simplify the performance evaluation of HICP. In that sense, we present the DAIs surveillance data from 5 adults medical-surgical ICUs at Buenos Aires, Argentina, members of Argentine Hospital Epidemiological Surveillance group since 1995.

Objective: To develop a simple benchmarking indicator that allows evaluating the HICP performance through DAIs rates in ICUs over time.

Methods: Data registered in specific infection control software (Epi-Control) from 5 tertiary care hospitals reporting DAIs from their ICUs from Jan 1995 to Dec 2008 was analyzed, all of them using the NNISS/NHSN methodology. The individual DAIs rates expressed as annual means were compared with NNIS/NHSN (US hospitals) and with International Nosocomial Infection Control Consortium (INICC: developing countries) 90^th percentile. A summary indicator was calculated as the proportion of total years of surveillance in which the annual means rate for a specific DAI was below the 90^th percentile for each benchmark.

Results: During a total of 58 surveillance years, (mean: 11.6; range 7-14) there were 2,027 DAIs for a total of 175,798 patients-days (overall infection rate: 11.5 events ‰ patients-days). Ventilator-associated pneumonia (VAP) was the most common infection (759; 37%; pooled mean: 11.4 ‰ device-days), followed by catheter-associated urinary tract infections (CAUTI) (690; 34%; pooled mean: 4.9 ‰ device-days) and central line-associated bloodstream infections (CLAB) (578; 29%; pooled mean: 5.7 ‰ device-days). Considering all DAIs together, an overall of 73.7% and 100% of the total annual means rates were below NNISS/NHSN and INICC 90^th percentile, respectively. Specific DAIs rates were maintained below the referred standards during a total time of 62.1% and 100% for VAP; 69.0% and 100% for CLAB, and 87.7% and 100% for CAUTI, respectively. When INICC 75^th percentile was considered, all DAIs rates were below the standard more than 90% of the time.

Conclusions: As these ICUs have been more than 90% of the time below the 75^th percentile benchmark of the developing countries, they should move towards a much stricter standard such as 90% of the time below 90^th percentile of US Hospital as part of quality improvement program. Otherwise, our data prove that even hospitals in the developing world can achieve low DAIs rates as in the developed one.