Value and price of ventilator-associated pneumonia surveillance as a quality indicator

ICU is growing but the methodology is still under development and subject to debate [1-3]. Ventilatorassociated pneumonia (VAP) seemed to come close to being an important, valid, reliable, responsive, interpretable and feasible outcome parameter [3]. We therefore decided to measure VAP incidence on a regular basis. A prospective study was carried out with yearly assessment of the incidence of VAP during a 3-month period. Defi nition of VAP was based on the recommendations of the Centers for Disease Control and Prevention [4]. Overall, out of 550 patients ventilated for >48 hours, only two cases of defi nite VAP were observed [5]. Because no further improvement could be achieved in this fi eld, we turned our attention to other outcome parameters. A perceived rise in incidence of VAP led us to repeat our evaluation, despite growing concern about the importance, validity and reliability of VAP as a quality indicator [2]. With the same methodology we measured the incidence of VAP again. Compared to our previous research, we observed a signifi cant (P < 0.001, chi-square test) increase in VAP in accordance with our impressions (Table 1). Th e incidence of VAP in our unit is still below that reported in the literature [2]. When used as a benchmark, we are performing well. However, when used as a quality indicator over time the results should lead to concern. Even if patients diagnosed with VAP do not have real VAP but colonization, atelectasis, or fl uid overload, these conditions are also detrimental for the patient and should be avoided [2]. As a benchmark, VAP incidence might have limited value [1,2]. Th is is mainly due to inappropriate case mix correction and to diagnostic inaccuracy. Fear of being judged on disputable quality indicators such as interhospital benchmarks is a serious threat to the probably valuable use of intra-hospital trend analysis of quality indicators. Used as a longitudinal quality indicator in a single centre, VAP is less threatened by case-mix diff erences and the limited sensitivity and specifi city of the VAP diagnostic criteria. Th e price of this quality assessment is considerable. Th e workload of this 14-week evaluation resulted in an estimated cost of 20,000 euros. In our view, measurement of VAP incidence has its value as an intra-hospital quality indicator but not as a benchmark.