Ventilator-Associated Pneumonia

Is There Any Gold in These Standards?

Standard: something established by authority, custom or general consent as a model or example.

Middle English, from Old French estandard rallying point.¹

Ventilator-associated pneumonia (VAP) is not only a common hospital-acquired infection but is said to be the most frequent infection in the ICU.² A literature search also suggests it must be one of the most studied and written-about topics in critical care. There seems to be a relatively constant risk per day of developing pneumonia while receiving mechanical ventilation.³ It is a problem that every hospital must address. Regardless how common it is, and despite the voluminous literature already in existence, controversies remain in every aspect of VAP including epidemiology, diagnosis, treatment, prevention, and outcomes assessment.

In this issue of CHEST (see page 2115), Rello et al have defined several risk factors and examined clinical and economic outcomes data. This is the largest US study of its kind to date. The authors are to be congratulated for tackling and dissecting the MediQual Profile database. Employing a vast database gives more credence to their conclusions about length of stay and added cost of care. They also openly discuss the limitations of their study, including difficulties regarding the definition of VAP and assessment of other risk factors. These limitations, however, are not unique to the study of VAP.

The challenges involved in the evaluation of VAP really serve as a paradigm for other ICU research issues. A recent study in CHEST looked at difficulties encountered in studies involving the critically ill.⁴ Although the study by Rello et al was a retrospective matched cohort study, and the recent CHEST study referred to the design of randomized clinical trials, the issues are similar. How do we define diseases and syndromes? How do we account for heterogeneous groups of patients undergoing multiple interventions? Can the measured outcomes truly discriminate risks and benefits? For example, though mortality rate may be the most common and easiest outcome to record, in the ICU it may be too insensitive to detect small benefits of certain interventions.⁴

Another study looked at several different definitions of VAP.⁵ Depending on criteria used, incidence of VAP varied from 4 to 48%. Furthermore, risk factors predicting VAP varied among the definitions. The particular basis for diagnosis may also account for differences in reported outcomes.⁶ Prior treatment with antibiotics may influence the predictive value of diagnostic techniques as well.

The Health and Science Policy Committee of the American College of Chest Physicians convened a group of experts to develop recommendations based on the literature.⁷ They suggested one of two management options. The first involved quantitative culture techniques, and the second used selection of antibiotics based on risk factors, local epidemiology, and resistance patterns. Both options are grade D, meaning no definite evidence or consensus opinion exists. This group also noted that the reported sensitivity of quantitative BAL and protected specimen brush (PSB) varied widely. These techniques are not well standardized, and there was no conclusive evidence or consensus indicating preference for one invasive test over another. Light⁸ suggested that BAL and PSB are simply variable dilutions of the endotracheal aspirate. He goes on to describe “the common sense notion that specimens obtained from locations only 5 to 15 cm apart along a widely patent airway in continuous motion are unlikely to have substantially different bacterial populations.” In any case, there is no high level evidence proving that any particular quantitative culture technique provides better clinical outcomes than empiric treatment.⁷

What should the clinician do when there is no “gold standard” approach to diagnosis, treatment, or outcomes? Rather than view it as a depressing controversy with no hope for resolution, I would suggest the study of VAP is ripe for further exploration. Study populations can be better defined and outcomes should be evidence based. When designing trials, researchers must decide philosophically if it is more important for the intervention to show benefit under ideal conditions or those encountered under usual practice circumstances.⁴ Despite the large volume of extant literature, many questions remain unanswered. Which patients benefit from continuous subglottic suctioning? Do certain surgical procedures predispose patients to VAP more than others? What is the relative risk for different types of trauma? I have wondered if the mode of mechanical ventilation itself, the inspiratory flow rate, or the inspiratory waveform in some way predisposes to or protects from the deposition of bacteria into the lower respiratory tract and thus affects the subsequent risk for VAP?

Finally, what should we at the bedside do today? I would argue for a practical and common sense approach. When possible, patients receiving ventilation should be positioned at a 45° head-up angle to decrease the risk for aspiration of gastric contents. Pick a definition of VAP for your institution and apply it consistently. Work closely with your hospital infection control committee to track cases. In hospitals with surveillance programs, nosocomial infections decreased almost a third over a 5-year period.⁹ Know your local antibiotic resistance patterns and apply this information when choosing empiric therapy and awaiting culture results. Review data with your medical staff, and when you make changes in ICU practices observe for differences in whatever outcome variable you choose.

Most editorials like this usually end with a plea for multicenter randomized prospective controlled trials. While this represents a laudable goal, it simply is not a feasible approach to answer every clinical question we encounter daily. Medicine has always been part art and part science. We have no reason to believe that the study of VAP should be any different. There remains an important role for clinical judgment. And … don’t forget to wash your hands!