Ventilator-Associated Pneumonia as a Quality Indicator for Patient Safety?

Ilker Uc kay 1 Qanta A. Ahmed 0 Hugo Sax 1 Didier Pittet () 1 0 Division of Pulmonary and Critical Care Medicine, Allergy and Clinical Immunology, Medical University of South Carolina , Charleston 1 Infection Control Program, University of Geneva Hospitals and Faculty of Medicine , Geneva , Switzerland The economic and clinical burden of ventilator-associated pneumonia (VAP) is uncontested. In many hospitals, VAP surveillance is conducted to identify outbreaks and to monitor infection rates. Here, we discuss the concept of benchmarking in health care as modeled on industry, and we contribute personal arguments against considering the VAP rate as a potential candidate for benchmarking or for monitoring the quality of patient care. Accurate benchmarking of VAP rates currently seems to be unfeasible, because the patient case mix is often too diverse and complicated to be adjusted for, and diagnostic criteria and surveillance protocols vary. Thus, the risk of drawing inaccurate comparisons is high. In contrast, some risk factors for VAP are modifiable and can be monitored and used as quality indicators. Process-oriented surveillance permits bypass of case-mix and diagnostic constraints. A well-defined interhospital surveillance system is necessary to prove that interventions on procedures do really lead to a reduction of VAP rates. - Nosocomial infections are common complications of a hospital stay [1]. Of these, ventilator-associated pneumonia (VAP) represents 5%18% of all infections [2, 3]. In a study involving 198 intensive care units (ICUs), the lung was the most common site of infection (68%) among patients with sepsis [4]. Overall, reported mortality rates for VAP have a range of 24%50% and can reach 76% in specific settings [5]. Although there are publications reporting no attributable mortality to VAP [5, 6], most authors believe that it contributes to 7%30% of additional mortality [79]. An additional 4 32 ventilator-days are ascribed to VAP [9, 10], and estimated attributable costs for 1 episode of VAP are reported to be as high as US $10,000 [11], US $16,000 [9], or even more [12]. At least 30% of all nosocomial infections are believed to be preventable [13]. Lowering the incidence of VAP would be an important quality improvement for patient safety. In a society where consumers consult restaurant ratings before making dinner reservations, the drive for hospital ratings is already palpable, and hospital leaders are more interested than ever in improving quality of care and in lowering costs [14]. In the United States, the Joint Commission [15] ratings propose specific patient safety targets before awarding accreditation, and this culture is already traveling overseas, with the establishment of the Joint Commission International Hospital Accreditation Process. But can VAP rates be used to draw conclusions? Moreover, is it fair to compare pneumonia rates across institutions? Is benchmarking meaningful, valuable, and, finally, to be recommended? BENCHMARKING: COPYING THE BEST Benchmarking was defined in 1989 by Camp as a continuous process of measuring products, services and practices against the toughest competitors or those companies recognized as industry leaders [16, p. 320]. In the early 1980s, the Xerox company found itself increasingly vulnerable to intensive competition from both US and Japanese companies, and its market share in copiers came down sharply from 86% in 1974 to only 17% in 1984. A leadership through quality policy was instigated with the revolutionary concept of benchmarking. Xerox looked first at internal company processes, followed by an assessment of its competitors, and collected data on key processes of best-practice companies. These critical processes were then analyzed to identify and define improvement [17]. To date, Xerox has conducted 1400 benchmarking studies and benchmarks itself against the best firms in every aspect of the market. The company now attributes 10% of annual productivity improvements to the lessons of benchmarking, and Xerox products are themselves once again industry benchmarks in certain product groups. Benchmarking has gained widespread acceptance in the private industry and is thought to lead to breakthrough improvements [18]. It serves to compare results, as well as structures and processes leading to these results. Two main types of benchmarking in private industry have evolved over time. First, internal benchmarking compares variations in differing units or departments within the same institution. Similar internal functions serve as pilot sites for conducting benchmarking through analysis of all processes involved in the task. A more covert internal benchmarking exists, which is the comparison of all the processes and policies of the same unit at different times. Second, competitive benchmarking is the study and measurement of ones policies against those of the best competitors. BENCHMARKING IN HEALTH CARE: BUSINESS AT THE BEDSIDE In health care settings, many attempts to benchmark have also been made on the basis of best practices [19], especially, but not exclusively, in the area of cardiovascular medicine [20]. Theoretically, every interhospital comparison adjusted for patient case mix can be considered to be benchmarking. Examples include the callto-needle time for thrombolysis in acute myocardial infarction [21], compliance with guidelines in the management of chronic heart failure [22], b-blocker prescription after myocardial infarction [23], treatment modalities for peripheral arterial disease [24], care for schizophrenia in the health care system [25], and the handling of recommendations for adolescent sexual health on the basis of comparisons with international best practices [26]. These were examples related to the process level of benchmarking. Examples of outcome benchmarking would be the assessment of the mortality of patients with diabetes; questioning the quality of diabetes care [27]; assessment of mortality of myocardial infarction related to b-blocker use [28]; assessment of survival after coronary artery bypass grafting, according to hospital-procedure volume [29]; and management of postoperative pain [30]. In the case of VAP, possible scopes of benchmarking could be the reduction of VAP risk by analysis and comparison of risk factors, prevention by comparison of preventive measures, comparison of clinical and/or microbiological trends, comparison of true incidence rates, and treatment and choice of antibiotics, with their impact on outcome. Other potential advantages would be the development of improved surveillance systems to follow emerging trends, catalyze action, activate administrative support, motivate health care staff, and acquire positive public and media attention in the overall context of the current trend for hospital ratings and public reporting. But is this really possible, and what are the difficulties to be encountered along the way? VAP AS A CANDIDATE FOR BENCHMARKIN (...truncated)