Pulmonary Infiltrates in The ICU - Management (Method of Antoni Torres MD) 

In our intensive care unit, we have developed an approach to diagnosis and treatment of ICU pneumonia (Figure 1). The major differential diagnosis of pulmonary infiltrates (print separately) includes only two entities in which antibiotics are clearly indicated: pneumonia and aspiration (Table 2). The presence of an infiltrate on simple chest x-ray raises the possibility that a patient may have pneumonia.  Our suspicion of hospital-acquired pneumonia occurs when two out of three clinical criteria are noted in a patient with pulmonary infiltrate: fever or hypothermia, leukocytosis or leukopenia and purulent respiratory secretions.

Table 2. Differential Diagnosis for Pulmonary Infiltrates in ICU Patients

Pneumonia Aspiration Atelectasis  Pulmonary edema Cardiogenic Noncardiogenic Pleural effusion Hemorrhage  Drug-induced

 

Figure 1:  Clinical Suspicion of Hospital-Acquired Pneumonia

APPROACH

               We calculate the Clinical Pulmonary Infection Score (CPIS) which allows objective analysis of the varying clinical and radiologic variables (Table 1). Cultures of the lower respiratory secretions are immediately taken (Table 3). We emphasize the use of the Gram stain of the sample of respiratory secretions; this improves diagnostic value of the score. Culturing should be performed prior to initiation of antibiotic treatment, but it should not delay the administration of antibiotics.

               Collection may be performed by sputum, tracheobronchial aspirate, bronchoalveolar lavage or protected specimen brush. Blood cultures should be obtained.  Pleural fluid cultures should be obtained, if the effusion is large and the cause is uncertain or if etiologic diagnosis of pneumonia has eluded identification.  Legionella urinary antigen is performed routinely given its occurrence as an occult cause of hospital-acquired pneumonia.  If the patient has been only recently hospitalized within 5 days, a urinary antigen for Streptococcus pneumoniae is performed.

               We obtain C-reactive protein (CRP) and procalcitonin (PCT). Elevated levels suggest more severe pneumonia.

Table 1. Clinical Pulmonary Infection Score

Score	Day 0		Day 3	Score
	Temperature, ºC        ³38.5º - 38.9º = 1 point               ³39.0º - 36.0º = 2 points		Temperature, ºC         38.5º - 38.9º = 1 point                39.0º - 36.0º = 2 points
	Blood leucocytes, mm-3        <4.000 or >11.000 = 1 point          50% band forms = add 1 point		Blood leucocytes, mm-3         <4.000 or >11.000=1 point          50% band forms = add 1 point
	Tracheal secretions           Presence of non-purulent tracheal secretions = 1 point            Presence of purulent tracheal secretions = 2 points		Tracheal secretions           Presence of non-purulent tracheal secretions = 1 point           Presence of purulent tracheal secretions = 2 points
	Oxygenation: PaO2/FIO2             >240 or ARDS = 0 point             < 240 and no ARDS = 2 points		Oxygenation: PaO2/FIO2          >240 or ARDS = 0 point          < 240 and no ARDS = 2 points
	Pulmonary radiography           No infiltrate = 0 point           Diffuse or patchy infiltrate = 1 point           Localized infiltrate= 2 points		Pulmonary radiography         No infiltrate = 0 point         Diffuse or patchy infiltrate = 1 point         Localized infiltrate= 2 points
	Microbiological Data           Pathogenic bacterial cultured in rare or hight quantity or no growth = 0 point           Pathogenic bacterial cultured in moderate or heavy quantity = 1 point           Same pathogenic bacterial seen             on Gram stain = add 1 point		Microbiological Data           Pathogenic bacterial cultured in rare or hight quantity or no growth = 0 point           Pathogenic bacterial cultured in moderate or heavy quantity = 1 point           Same pathogenic bacterial           seen on Gram stain = add 1              point

Total Day #0 = _________                                        Total Day #3 = _________

Table 3. Diagnostic Tests in the Workup of Pulmonary Infiltrates in the ICU

1.  Respiratory Cultures Sputum Endotracheal or Tracheobronchial aspirate** [Toni: should this be transbronchial?] Bronchoalveolar lavage (BAL) or mini-BAL ** Protected brush specimen (PBS) **   2. Two blood cultures   3. Pleural fluid culture if parapneumonic effusion.   4. Legionella pneumophila urinary antigen and Streptococcus pneumoniae urinary antigen.   5. CBC, electrolytes, hepatic and renal function tests.   6. Arterial blood gases   7.  C-reactive protein (CRP) and procalcitonin

Send samples to microbiology laboratory immediately (or if not possible, refrigerate at 4ºC for a maximum of one hour) Gram staining, intracellular organism counting (only in BAL and mini-BAL) and quantitative cultures *** should be done.

*Collection of cultures should not delay the initiation of empiric treatment in patients with severe sepsis.

**These techniques may be performed by bronchoscopy by blinded procedures.

***Quantitative cultures are performed with the respiratory secretions obtained by transbronchial aspirate, BAL or PBS. The cut-off points for determining infection colonization are: Transbronchial aspirate 10⁵CFU/mL; BAL 10⁴ CFU/mL and PSB 10³ CFU/mL.

Assessment of CPIS, Clinical Status, Gram Stain 

               If the CPIS < 6, systemic inflammatory response syndromes (SIRS) (Table 9) is not present and the Gram stain of respiratory secretions is negative or intracellular organisms < 2%, then pneumonia is not likely. Patients fulfilling all of these criteria need not receive antibiotic treatment but be strictly monitored.  This is our unique addition to the protocol designed by Singh et al.; we have found it to be reliable and clinically-plausible. These patients do not fulfill classic criteria for pneumonia and they are clinically stable.  Monitoring such patients does not pose a substantial risk for poor outcome, since antibiotics can be added if deterioration is observed.  Keep in mind, it is probable that noninfectious causes of pulmonary infiltrates will be diagnosed in such patients within 3 days.

               At the other extreme, if any of the following are present: CPIS > 6, presence of SIRS, and Gram stain of respiratory secretions shows a predominant bacterium or intracellular bacteria, these patients most likely have pneumonia.  Broad spectrum antibiotic treatment should be initiated (Figure 2).

               If the CPIS is > 6 and SIRS is not present, pneumonia is still likely, and broad spectrum antibiotic therapy is initiated immediately.

               If the CPIS is < 6 and the Gram stain of the repiratory tract sample is difficult to interpret, then we administer monotherapy.  In the original protocol by Singh, et al., a quinolone was used for monotherapy, and in an NIH proposed protocol, a carbapenem was used.

Figure 2: Algorithm for the Management of Patients with Pulmonary Infiltrates

Table 8. Definitions

A. SIRS: 2 or more of the following variables: 1.      Fever >38◦C or < 36◦C 2.      Heart rate >90 beats per minute 3.      Respiratory rate >20 breaths per minute or PaCO2 <32 mm Hg 4.      Abnormal white blood cell count (>12,000/mm3 or <4,000/ mm3 or >10% bands) B. Bacteremia: bacteria within the blood stream (does not always lead to SIRS or sepsis)  C. Sepsis: SIRS plus a documented or presumed infection. D. Severe sepsis: aforementioned sepsis criteria with associated organ dysfunction, hypoperfusion or hypotension.  E. Sepsis induced hypotension: presence of a systolic BP <90 mmHg or a reduction of > 40 mmHg from baseline in the absence of other causes of hypotension.”  F. Septic shock: Persistent hypotension and perfusion abnormalities despite adequate fluid resuscitation.  G. Multiorgan dysfunction syndrome: state of physiological derangements in which organ function is not capable of maintaining homeostasis.

CPIS < 6

               If the CPIS < 6, and the Gram stain of the respiratory tract specimen is difficult to interpret, antibiotic therapy can be limited. Whether the CPIS is a good predictor of pneumonia is controversial and numerous studies have questioned its sensitivity and specificity. However, in our approach, the CPIS is not used as a proven aid to diagnose ICU pneumonia; instead, its use is to select those patients who do not need intensive and prolonged broad spectrum antibiotic therapy. The usage for this purpose has been validated in one study. So, for these patients, we administer antibiotic monotherapy initially. Since the classical signs of pneumonia are largely absent and since the patient is clinically stable, most clinicians would agree that limited antibiotic therapy (monotherapy) of short duration (3 days) is prudent until results of cultures are known. Studies estimate that for ICU patients with pulmonary infiltrates 70%-80% do not have pneumonia, but currently most will receive combination broad spectrum empiric antibiotic therapy with duration from 5-14 days. Receipt of unnecessary antibiotics in patients without confirmed pneumonia is linked to higher mortality.

               The major reason contributing to “spiraling empiricism” in antibiotic use in the ICU is that physicians are unwilling to risk missing a treatable infection. Because ICU pneumonia carries substantial mortality, administration of broad-spectrum antibiotics to most patients with pulmonary infiltrates has emerged as the predominant practice in the ICU. The use of the CPIS shown in Figure 2 allows the physician flexibility in initially managing patients with a perceived treatable infection. However, the number of antibiotics to be given is limited (monotherapy) as is the duration (3 days). Such an approach has led to decreased costs for antibiotics and minimized the emergence of antimicrobial resistance. As importantly, outcome was more favorable when compared to a control group of patients with CPIS < 6 who received broad spectrum antibiotic therapy for 5-21 days.

Re-Evaluation on Day 3

               On the third day, response to empiric antibiotics is observed using clinical, laboratory and radiologic parameters and the re-calculated CPIS (Figure 3).  The criteria of treatment failure are: 1) absence of improvement in the PaO₂/FiO₂ or the need for mechanical ventilation, 2) persistence of fever or hypothermia, 3) 50 % progression in the pulmonary infiltrates, and 4) development of septic shock or multiple organ dysfunction syndrome. For patients whose original CPIS < 6, if the patient appears stable and the cultures are negative, all antibiotic treatment should be discontinued on day 3.  For those patients in whom cultures are revealing, antibiotic therapy should be focused on the specific pathogen and the spectrum narrowed. In patients with increasing level of biomarkers (RCP and procalcitonin) on day 3-5, an extensive microbiological and radiological re-evaluation is warranted and broadening the spectrum of the antimicrobial therapy is done.

Figure 3:  Follow up of Patients with Pulmonary Infiltrates

Antibiotic Selection

               Empiric antibiotic therapy (print separately) should be based on local in vitro susceptibility patterns and patient risk factors for infection by multiresistant microorganisms (Table 4). For patients with CPIS > 6, and risk factors for infection by multiresistant microorganisms or duration of mechanical ventilation more than 5 days, we use a combination of an anti-pseudomonal beta-lactam antibiotic with the addition of a quinolone or an aminoglycoside (Table 5, Figure 6). This is a reasonable choice for both Pseudomonas aeruginosa or multi-drug resistant microorganisms. 

Table 4. Risk Factors for Infection by Multiresistant Microorganisms

Risk factors for Multi-Resistant Microorganisms

Antibiotic treatment within the last 90 days (> 5 days) Current hospital admission or within the last 90 days (> 5 days) Immunosuppressive disease and/or treatment Chronic dialysis within the last 30 days Epidemic outbreak in the unit by multiresistant organism

Table 5: Initial Empiric Antibiotic Treatment For Hospital-Acquired Pneumonia or VAP Of Late Onset Or In Patients With Risk Factors For Resistance

 

Microbial Etiology	Combination antibiotic treatment
Microorganisms from Table 3 plus:   Pseudomonas aeruginosa Klebsiella pneumoniae (ESBL+) Serratia marcescens Acinetobacter spp. Other nonfermentative GNB MRSA Legionella pneumophila	Antipseudomonal cephalosporin (ceftazidime or cefepime)*                     or Carbapenem (imipenem, meropenem)*                     or Beta-lactam / betalactamase inhibitor (piperacillin / tazobactam)*                      + Antipseudomonal quinolone (ciprofloxacin, levofloxacin)**                      or Aminoglycoside ** (amikacin)                      ±       Linezolid or vancomycin***

GNB = Gram negative bacilli

ESBL = Extended-spectrum beta-lactamase

MRSA = Methicillin-resistant Staphylococcus aureus

*The choice of beta-lactam antibiotic is made as follows: patients who have not received any antipseudomonal beta-lactam antibiotic within the last 30 days are administered piperacillin/tazobactam or antipseudomonal beta-lactam cephalosporin. Patients who have received these prior antipseudomonal drugs are given a carbapenem. Patients with infection by ESBL-producing microorganisms are treated with carbapenem regardless of the results of the antibiogram.

**For combination empiric therapy for multiresistant gram-negative bacilli, an antipseudomonal quinolone is used in cases of renal failure or concomitant use of vancomycin. In other settings combined empiric therapy is initiated with amikacin and is maintained for a 5 day period. 

***Empiric therapy aimed at MRSA is initiated in patients with established colonization, previous infection by MRSA, or implementation of mechanical ventilation for more than 6 days. Our antibiotic of choice is vancomycin. However, linezolid is used in patients allergic to vancomycin, creatinine values ≥ 1.6 mg/dL or in patients presenting signs of infection after 48 hours of vancomycin therapy and in those in whom MRSA has been isolated.

               Empiric therapy for methicillin-resistant Staphylococcus aureus (MRSA) is implemented only in patients who demonstrate colonization or previous infection by MRSA or who have received mechanical ventilation for more than 6 days. In many hospitals, linezolid is preferred to vancomycin.

               In the absence of risk factors for infection by multiresistant microorganisms or in patients who have been hospitalized or have received mechanical ventilation for less than 5 days, pathogens of community-acquired pneumonia should be covered; Ceftriaxone or levofloxacin can be added for these pathogens (Streptococcus pneumoniae, Legionella pneumophila) (Table 6, Figure 5, 6).

Table 6. Initial Empiric Antibiotic Treatment In Hospital-Acquired Pneumonia or VAP of

Early Onset In Patients Without Risk Factors For Resistance

Probable Microorganism	Empiric Antibiotic
Streptococcus pneumonia   Haemophilus influenzae   Enteric Gram-negative bacilli Escherichia coli Klebsiella pneumoniae Enterobacter spp. Proteus spp.	Ceftriaxone                                or                        Levofloxacin

 

Figure 5:  Empiric antibiotic treatment of VAP in patients without factors of infection by P. aeruginosa

Figure 6: Empiric antibiotic treatment of VAP in setting of risk of infection by P. aeruginosa or

   by multiresistant microorganisms

 

               Table 7 shows the antibiotics, the doses, treatment schedule and the length of infusion.  If beta-lactam agents are selected, administration by continuous infusion should be considered.

Table 7. Antibiotic Dosages and Timing

Antibiotic	Doses	Interval of administration	Infusion time
Ceftriaxone	1 g	12 hours	1 / 2 - 1 hour†
Levofloxacin	750 mg	12 hours*	1 / 2 hour
Ceftazidime	2 g	8 hours	2 - 3 hours†
Cefepime	2 g	8 hours	2 - 3 hours†
Imipenem	0.5 g	6 hours	1 hour†
Meropenem	0.5 – 1 g	6 hours	2 - 3 hours†
Piperacillin/Tazobactam	4 / 0.5 g	6 hours	2 - 3 hours†
Ciprofloxacin	400 mg	8 hours	1 / 2 hour
Amikacin	15 mg / Kg	24 hours **	1 / 2 - 1 hour
Vancomycin	1 g	8-12 hours***	1- 3 hours*
Linezolid	600 mg	12 hours	1 hour

*Administer this dose for 3 days and after continue with 500 mg / 24 hours

**Adjust the dosage according to PK / PD parameters

***Initiate this dose with 24 hours, measure trough blood levels prior to the following dosage and adjust the levels according to values.

†For beta-lactam agents and vancomycin, continuous infusion should be considered.