microbiologic specimens is beyond the scope of this brief chapter, the following should be
mentioned:
l Suspected line sepsis: A decision must be made whether to remove one or more vascular
access devices or to rely on clinical observation combined with “through-the-line”
blood cultures obtained simultaneously with blood cultures drawn by venepuncture.
l Suspected ventilator-associated pneumonia: Data based on specimens obtained by
bronchoscopy, using either bronchoalveolar lavage or bronchial brushing, have
added enormously to our understanding and treatment of VAP. Whether such
specimens should be part and parcel of routine ICU practice remains controversial.
l The obtunded patient: One should remember the possibility of meningitis and/or
encephalitis, and the old adage “if you think of a lumbar puncture then do one” still
remains true.
l Blood cultures: Scrupulous collection technique is required especially to avoid
unnecessary treatment of contaminating microorganisms, most commonly coagu-
lase-negative staphylococci (usually, MRSE). Through-the-line cultures are to be
discouraged except for diagnosis of line sepsis, as mentioned above. At least two
cultures should be obtained.Pretreatment cultures provide much of the basis for subsequent simplification.
In 1977, Lowell Young and his colleagues proposed “the rules of three” for bloodstream
infections (21). They pointed out that if three blood cultures have been obtained and that if at
the end of all three days these specimens remain sterile, it becomes progressively unlikely that
bloodstream infection will be documented by those specimens. This rule takes advantage of
the relatively rapid isolation of most aerobic pathogens. With only rare exceptions, such as the
“HACEK” organisms (certain fastidious gram-negative rods that occasionally cause infective
endocarditis) andBrucellaspp., this rule applies to most organisms likely to be encountered in
the ICU, including yeasts. Numerous studies confirm this clinical insight. Indeed, one can
argue that improvements in microbiologic techniques now mandate a revision to “the rules of
two.” One could make a case for “a rule of one,” and it is certainly conceivable that, at some
point during the 21st century, molecular techniques will make it possible to rule in or out
various pathogens within a matter of minutes.
De-escalation therapy has been best studied in the case of VAP. VAP, discussed at length
elsewhere in this volume, constitutes the single-most common cause of death from hospital-
acquired infection. Serial studies of respiratory secretions from patients on ventilators
commonly reveal an all-too-familiar “parade of pathogens” whereby increasingly difficult-to-
treat bacteria emerge during therapy, prompting “spiraling empiricism” in the use of
increasingly broad-spectrum and potentially toxic agents. For effecting what amounts to a
revolution in our approach to VAP, due credit must be given to the French workers who
championed the use of bronchoscopy to obtain specimens for bronchoalveolar lavage (BAL) or
the protected specimen brush (22).
Mention will be made here of two studies from the substantial and growing literature on
de-escalation therapy for VAP, based, at least in part, on specimens obtained by bronchoscopy.
Singh and colleagues conducted a study whereby patients with less extensive evidence of
pulmonary infection were randomized to receive standard care (antibiotics for 10–21 days) or
to be reevaluated after three days. Patients who were reevaluated at three days experienced
similar mortality but were less likely to develop colonization or superinfection by resistant
organisms (15% vs. 35%,p= 0.017) (23). Rello and colleagues made a practice of reevaluating
patients after two days of therapy, taking into account clinical improvement and culture
results. Approximately 40% of their 115 patients were on a trauma service. More than one-half
(56%) of their patients had their therapy modified, and the ICU mortality rate was significantly
lower (18% vs. 43%,p<0.05) in patients whose therapy was modified (24).
The concept of de-escalation and also of limiting the duration of antibiotic therapy to
seven or eight days for uncomplicated VAP (and other HCAPs) has now been endorsed by the
American Thoracic Society (13). Current and future investigators will no doubt take advantage
of evolving diagnostic techniques to refine and extend these recommendations to most, if not
all, ICU infections.
492 Ahuja et al.