Pneumonia is a common disease and a leading cause of death. Indeed, William Osler recognized that most old people die from pneumonia rather than from old age itself [1]. Pneumonia is therefore a leading cause of hospitalization and amongst the most common reasons for admission to the ICU. Indeed, over 50 % of patients with severe sepsis have pneumonia. Most bacterial pneumonias, with the notable exception of Legionella Pneumonia, Tuberculosis and Leptospiral Pneumonia are caused by the aspiration of oropharyngeal material colonized with the causative pathogen. This implies that the “causative” pathogen together with low-virulence oral flora are aspirated in most pneumonia. Appropriate initial antimicrobial therapy, defined as the use of at least one antibiotic active in vitro against the causative bacteria, is associated with reduced length of hospitalization, reduced morbidity and reduced mortality when compared with patients receiving initial inappropriate therapy [2–4]. As inappropriate initial antibiotic therapy is usually associated with infection with a drug-resistant pathogen (DRP), and as DRPs are usually acquired in specific settings, the treatment of pneumonia has traditionally been based on the setting in which the pneumonia develops. The number of individuals receiving health care outside the hospital setting, including home wound care and infusion therapy, dialysis, nursing homes and long-term acute care (LTAC) facilities is increasing. One of the most frequent causes of hospitalization and mortality in these patients is pneumonia. Hence a new class of pneumonia was identified; healthcare associated pneumonia (HCAP). Consequently, the 2005 and 2007 guidelines for the management of pneumonia by the American Thoracic Society and the Infectious Diseases Society of America recommend that pneumonia should be classified into one of three categories at diagnosis: (1) community-acquired pneumonia (CAP), (2) healthcare-associated pneumonia (HCAP), and (3) hospital-acquired pneumonia (HAP) [5, 6]. These three categories of pneumonias are further subdivided into the following “word-soup” subtypes according to the setting where the pneumonia develops, the severity of illness and the risk of aspiration, namely:

The mortality of patients with HCAP is higher than that of patients with CAP (about 30 % vs 5 %); this is likely related to the comorbidities and poor-functional status of patients with HCAP rather than the virulence of the pathogen or the prescription of initial inappropriate antibiotic therapy [7–9].

It has been argued that patients with HCAP and HAP are at a higher risk of infection with DRP’s such as Pseudomonas aeruginosa and methicillin–resistant Staphylococci (MRSA) and therefore require an initial broad spectrum multidrug antibiotic regimen. It has been assumed that the three major categories of pneumonia are distinct clinical entities that require specific antimicrobial regimens. However, increasing evidence suggests that there is much overlap between these conditions and that a more unified approach to the treatment is more practical. In essence these pneumonias differ by the spectrum of colonizing organisms which are aspirated and it may therefore be more useful to risk stratify patients according to the likelihood of developing infection with gram-negative and DRPs. The traditional approach assumes that patients with HCAP are infected with DRPs while CAP patients are infected with multi-sensitive organisms (MSO). However, a small percentage of patients with CAP are infected with DRPs while a significant percentage of patients with HCAP are infected with MSO (see Fig. 17.1). Indeed, only 10–30 % of patients with HCAP are infected with DRPs [7, 10, 11]. A more useful schema is to assess the risk of infection with a pathogen not susceptible to ceftriaxone, ampicillin-sulbactam, macrolide or a respiratory fluoroquinolone; these known as CAP drug-resistant pathogens (CAP-DRPs) [8]. In a large prospective observational study Shindo et al. reported infection with CAP-DRPs in 26.6 % of patients with HCAP and 8.6 % of patients with CAP. Remarkably, the risk factors for CAP-DRPs were almost identical in patients with CAP and HCAP. Based on this concept, the following have been identified as risk factors for infection with a CAP-DRPs in both community and healthcare associated pneumonia [8, 10, 12]:

The risk of a CAP-DRPs increases with the number of risk factors being about 3 % with no risk factors, 10 % with one risk factor and greater than 80 % for 5 or more risk factors [8]. According to this revised treatment approach a patient with pneumonia and one or more risk factors for a CAP-DRP (irrespective of where the patient develops the pneumonia) should receive broad spectrum antibiotics, while all others should be treated with narrow spectrum antibiotics (traditional CAP antibiotics) [8, 9]. It may be also reasonable to tailor the “aggressiveness” of the CAP-DRPs regimen according to the number of risk factors for a CAP-DRP. Poor functional status (Barthel Index score <50) is a major factor increasing the risk of infection with a CAP-DRP; these patients usually have multiple other risk factors for infection with a CAP-DRP [9, 13, 14]. The antibiotic regimen for those likely to be infected with a CAP multi-sensitive organism (CAP-MSO) is further stratified according to severity of illness (see algorithm below). In patients risk stratified to treatment with a CAP-DRP regimen, monotherapy is adequate once a known pathogen is identified. This strategy of initiating broad spectrum cover with two or more antibiotics and then narrowing the spectrum to a single agent when a pathogen is identified is known as “antimicrobial de-escalation” [15]. The de-escalation approach has been demonstrated to be associated with a reduction in mortality [16]. The exception to this “monotherapy” rule is patients with “severe-CAP” in whom combination therapy with ceftriaxone and a macrolide is recommended (see below)