Chest ACCP Member Benefits
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (43)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Campbell, S. G.
Right arrow Articles by Ackroyd-Stolarz, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Campbell, S. G.
Right arrow Articles by Ackroyd-Stolarz, S.
(Chest. 2003;123:1142-1150.)
© 2003 American College of Chest Physicians

The Contribution of Blood Cultures to the Clinical Management of Adult Patients Admitted to the Hospital With Community-Acquired Pneumonia*

A Prospective Observational Study

Samuel G. Campbell, MBBCh; Thomas J. Marrie, MD; Rosemary Anstey, ART, BAS; Garth Dickinson, MD and Stacy Ackroyd-Stolarz, MSc; for the capitaL Study Investigators{dagger}

* From the Department of Emergency Medicine (Dr. Campbell and Ms. Ackroyd-Stolarz), Dalhousie University, Halifax, NS; Division of Medicine (Dr. Marrie), University of Alberta, Edmonton, AL; Endpoint Research Ltd (Ms. Anstey), Mississauga, ON; and Division of Emergency Medicine (Dr. Dickinson), University of Ottawa, Ottawa, ON, Canada. {dagger} A list of participants and participating centers is given in the Appendix.

Correspondence to: Samuel G. Campbell, MBBCh, Department of Emergency Medicine, Queen Elizabeth II Health Sciences Centre, 1796 Summer St, Halifax, NS, Canada B3H 3A7; e-mail: sgcampbe{at}is.dal.ca


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
 Appendix
 References
 
Study objective: To assess the clinical usefulness of blood cultures (BCs) in the management of patients hospitalized with community-acquired pneumonia (CAP).

Design: A prospective, observational study to investigate the contribution of BCs to the management and outcomes of adult patients presenting with CAP. Setting: Nineteen Canadian hospitals.

Patients: Adults admitted to the hospital with CAP between January 1, 1998, and July 31, 1998. Interventions: The courses of therapy in patients for whom BC results yielded organisms considered to be clinically significant were analyzed to determine whether the BCs had contributed to management or outcome.

Measurements and results: Forty-three of 760 patients had significantly positive BC results. Patients with CAP who had BCs performed had a 1.97% chance (15 of 760 patients) of having a change of therapy directed by BC results. Patients in whom BCs yielded positive results had a 34.8% chance (15 of 43 patients) of having a change in therapy determined by BC results, and had a 58.1% chance (25 of 43 patients) of having a course of therapy contraindicated by BC results. Severity of illness, as measured by the pneumonia severity index, correlated poorly with the yield of BCs. BC results were positive in 8.0% of patients in risk classes I and II, 6.2% of patients in risk class III, 4.6% of patients in risk class IV, and 5.2% of patients in risk class V.

Conclusion: BCs have limited usefulness in the routine management of patients admitted to the hospital with uncomplicated CAP.

Key Words: blood cultures • clinical practice guidelines • community-acquired pneumonia • microbiology


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Conclusion
 Appendix
 References
 
Blood cultures (BCs) are considered a standard of care as part of the initial laboratory workup for patients admitted to hospital for communityacquired pneumonia (CAP) at most institutions in North America and, to our knowledge, are included in all major contemporary guidelines for CAP management.1 2 3 4 5 6 The clinical value of routine BCs has been questioned.7 8 9 10 11 12 13 14 15 16 In some instances, this questioning has been followed by impassioned pleas not to abandon routine BCs, with authors citing anecdotes of cases in which BCs proved invaluable to patient care.17 18 We sought to evaluate the contribution of BCs to patient management in a large number of Canadian patients admitted to the hospital with CAP.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Conclusion
 Appendix
 References
 
The capitaL Study was a multicenter, controlled clinical trial with cluster randomization to determine if the use of a clinical pathway improved the efficacy of treatment for CAP without compromising patient well-being. Participating hospitals were assigned to either implement a clinical pathway being tested (n = 9), which consisted of a clinical prediction rule to guide the decision regarding the site of treatment of CAP, levofloxacin therapy, and practice guidelines, or to continue conventional management of CAP (n = 10).

We predicted that data from the study would have the potential to demonstrate the utility of BCs in patients admitted with CAP. We conducted a prospective, observational study using the study population to investigate how the performance of BCs would affect the management and outcomes of adult patients admitted to hospital with CAP.

At the intervention sites, the study protocol called for BCs prior to starting antibiotic therapy. At the conventional therapy sites, BCs were performed at the discretion of the attending physicians. At the intervention sites, two BCs (one from each study arm) were recommended. BCs were processed at the microbiology laboratory at the participating sites according to the protocol of each laboratory.

BC results and other clinical information were sent from participating centers to a central data center. Data from patients with positive BC results were specifically analyzed by two investigators (S.G.C. and R.A.). Changes (or the lack of changes) in the antibiotic regimens of patients with positive BC results were examined in an attempt to assess the influence of the results on clinical management. For the purposes of this study, a change in management was defined as a change in antibiotic from, or the addition of another antibiotic to the empiric regimen. A step-down to an oral agent in the same class as the IV agent was not considered a change in management.

Patients
Eligible patients were adults presenting to the emergency departments of participating hospitals between January 1, 1998, and July 31, 1998, with at least two signs or symptoms of CAP (eg, temperature > 38°C, productive cough, chest pain, shortness of breath, crackles on auscultation) and whose chest radiograph showed an opacity compatible with the presence of acute pneumonia. Exclusion criteria included patients with an immune deficiency, in shock or requiring direct ICU admission, alcohol addiction, chronic renal failure, or pregnant or nursing women.19


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Conclusion
 Appendix
 References
 
A total of 2,804 patients with a suspected diagnosis of CAP were enrolled in the study. Of these, 1,061 patients were excluded for failure to meet inclusion criteria or for the presence of exclusion criteria. Of the remaining 1,743 patients, 716 patients were randomized to the intervention arm and 1,027 patients were randomized to the conventional arm (Table 1 ). One thousand twenty-two patients were admitted to the hospital on their first visit. Of these, BCs were drawn in 760 patients (74.4%), with a yield of 44 "significant" organisms (from 43 patients, 5.66%) [Table 2 ]. BCs that yielded organisms considered likely to be contaminants were excluded, unless isolated from multiple sites.


View this table:
[in this window]
[in a new window]

  		Table 1. Rate of Positive BC Results by PSI Class*

 

View this table:
[in this window]
[in a new window]

  		Table 2. Yield of BCs According to Study Arm

 
As was expected, the most prevalent pathogen isolated was Streptococcus pneumoniae, accounting for 68.1% of the isolates. Staphylococcus aureus accounted for 11.4% of the isolates and was tied with Escherichia coli as the second most common microorganism isolated from blood. One sample was positive for Haemophilus influenzae (2.3%) and for Enterobacteriaceae in seven isolates (16.3%): five E coli, one Klebsiella pneumoniae, and one Enterobacter agglomerans. There was one isolate of Enterococcus faecalis recovered from a patient who also had BC results positive for K pneumoniae. This was the only patient in whom polymicrobial bacteremia was identified (Table 3 ). Only one case of penicillin-resistant S pneumoniae (PRSP) was found. Fourteen organisms, all from the intervention arm, were specifically tested for levofloxacin sensitivity, and all were found to be susceptible: S pneumoniae (n = 12), S aureus (n = 1), and E coli (n = 1).


View this table:
[in this window]
[in a new window]

  		Table 3. Breakdown of Positive BC Results*

 
Antimicrobial Treatment Course in Patients With Positive BC Results
In patients for whom BC results were positive, continuation of the empiric regimen in light of the BC results was considered appropriate in three cases. In 23 cases (53.5%), the class of treatment was either not changed during the course of the illness, or was actually increased, despite organisms being cultured that were susceptible to a narrowerspectrum or cheaper antibiotic than the empiric regimen. A change to broader-spectrum coverage occurred in five cases of penicillin-susceptible organisms (four cases of penicillin-susceptible S pneumoniae [PSSP] and one case of S aureus), and in one case of E coli, suggesting that the physicians did not feel reassured by the culture findings in patients who were deteriorating (two patients subsequently died). In 17 cases (39.5%), the empiric regimen continued in spite of a BC result suggesting that it might be safe to step down to more specific, less expensive medication (Table 4 ).


View this table:
[in this window]
[in a new window]

  		Table 4. Sequence of Antibiotic Administration in Patients With Positive BC Results*

 
Of those cases in which treatment was changed, 14 cases had a less expensive regimen chosen (32.6%, or 1.8% of the total BCs drawn), although in one of PSSP cases, the change from erythromycin to clindamycin was unlikely to have resulted from the BC result. In one case of S aureus, IV treatment was appropriately changed to cloxacillin; however, the oral step-down regimen prescribed was amoxicillin, in spite of the demonstration of penicillin resistance and erythromycin sensitivity.

In only three cases (6.98% of positive results, 0.39% of BCs drawn, one each of methicillinsusceptible S aureus, PRSP, and methicillin-resistant S aureus [MRSA]) were regimens with more reliable coverage chosen, representing 0.4% of the total number of BCs drawn. The patient with MRSA died within 48 h of hospital admission. Treatment was changed on the first day, before the MRSA result would have been available. The patient with PRSP was discharged after 9 days, and the patient from whom S aureus was cultured was discharged after a 42-day hospital stay.

In 11 patients with positive BC results, antibiotics were administered IV throughout the course of treatment. Five of these patients died, two after having anaerobic coverage added to their regimen, and both of whom had penicillin-sensitive organisms (S aureus and S pneumoniae) grown on BC. Of the other patients who died during that hospital admission, one patient with MRSA had treatment changed from levofloxacin to cefotaxime, one patient had no change in agent despite the growth of PSSP on BC, and one patient with S pneumoniae was changed to penicillin G. Of the survivors treated only with IV antibiotics, two patients were changed from an empiric regimen that was not believed to adequately cover the organism cultured (to vancomycin in a patient with PRSP, and to cefazolin in a patient with S aureus). One patient with BC growth of PSSP continued on the empiric regimen, and another patient had broader-spectrum coverage added in spite of the BC result suggesting that in each, a narrower spectrum of treatment may have been appropriate. In one case of S aureus, empiric treatment was continued, again despite a BC result suggesting that a less expensive medication might have been appropriate. In one case in which E coli was cultured, a narrower-spectrum antibiotic was chosen.

Thirty-two of the patients with positive BC results had their antibiotic therapies stepped down from an IV to an oral agent during the course of treatment. Of these patients, the same or a similar class of agent was chosen in 17 cases. Of these 17 cases, 11 patients had BC results suggesting that it would have been safe to step down to oral penicillin; in 3 cases, BC results positive for E coli suggested that a less expensive regimen was indicated; and in 3 cases, the regimen continued in agreement with the suggestion of the BC result. Ten patients were appropriately stepped down to less expensive and narrowerspectrum drugs according to the BC result. One patient with PSSP was stepped down to a more expensive class of drug in contradiction to the BC result (cefuroxime to amoxicillin-clavulinate), and two patients with PSSP had clindamycin added (presumably on suspicion of aspiration), both in contradiction with the BC result. In one of the latter cases, clindamycin was included in the step-down regimen. In one patient with E coli, cefazolin was added; and, as mentioned above, one patient with positive BC results for S aureus was stepped down to amoxicillin, again both in contradiction with the BC finding (Tables 5 , 6 ).


View this table:
[in this window]
[in a new window]

  		Table 5. Relationship Between BC Result and Course of Treatment*

 

View this table:
[in this window]
[in a new window]

  		Table 6. Step-Down Therapy Following BC Result

 
In the intervention arm, levofloxacin was the drug specified on the pathway and, as the study drug, was provided at no cost to the institutions involved. To investigate whether this might have lead us to underestimate physicians’ willingness to change to less expensive medication, the incidence of appropriate step-down was compared between the two arms of the study (Table 6) . There was no statistical difference between study arms regarding step-down to a less expensive or narrower-spectrum therapy, considered to be in agreement with the BC result (p = 0.7) [Table 6 ].

Correlation of BC Positivity Rate With Pneumonia Severity Index Scores
In the patients in whom BC results were positive (Table 4) , 11 patients had pneumonia severity index (PSI) scores (described by Fine et al20 ) calculated as class I or II (PSI score < 71), 10 patients were in class III (PSI score of 71 to 90), 16 patients were in class IV (PSI score of 91 to 130), and 6 patients were in class V (PSI score > 130) [Table 1 ]. Calculated PSI scores for the five patients who died were all class IV (three patients) or class V (two patients).

Cost of Performing BCs
Calculated according to the cost of BCs at our institution ($41.70 per two sets of BCs per patient [Nova Scotia Department of Health rate for aerobic BC]),21 the BCs in our study population would have cost $31,281.60 (Canadian), or $2,085.44 per positive finding leading to a change in treatment. To calculate a crude estimate of the cost savings realized by the change to less expensive regimens, the cost of continuing the empiric regimen for the same length of time that the new choice was given was calculated, and the cost of the new choice was subtracted (ie, $3,250.92 - $792.71). In calculating costs, it was presumed that the same periods of IV and oral antibiotic administration would have been used had the empiric regimen been continued. The cost of savings attributed to BC-guided medication changes was calculated to be $2,458.21, or $3.23 saved per patient in whom BCs were ordered, a net cost to the system of $28,823.39, or $37.93 per BC. Each change in treatment attributed to BC results cost the system $1,921.56.


   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Conclusion
 Appendix
 References
 
This study is the largest to specifically and prospectively evaluate the clinical usefulness of BCs in CAP. Patients were enrolled from 19 different Canadian hospitals, adding to the generalizability of the findings. The yield of BCs in our study was 5.7%, which is similar to the 6 to 10% reported in most series.6

The proportion of organisms identified is also consistent with that reported in other studies, in that S pneumoniae represented more than two thirds of the positive BC results.6 22 23 E coli (at 14% of the total BCs) was the second most common organism cultured; this is similar to the findings of Marrie et al24 and Marston et al,25 who found E coli to be the second most common cause of bacteremic pneumonia in their studies (11.3% and 8.7% of bacteremic patients in each study, respectively).

The Utility of BCs in Guiding Therapy
Patients presenting with CAP who had BCs performed had a 1.97% chance (15 of 760 patients) of having a change of therapy directed by the BC result. In only three cases (0.4%) was this change likely to have improved the clinical outcome for the patient. Patients in whom BCs yielded positive results had only a 39% chance (15 of 43 patients) of having a change in therapy determined by the BC result, and had a 42% chance (18 of 43 patients) of having a course of therapy contradicted by BC results. These findings suggest that the physician’s clinical judgment is more influential in determining the course of medication than the results of microbiological investigations. The reticence of physicians to change to a narrower spectrum of coverage has been noted by numerous authors.12 16 26 27

Our findings are similar to those of previous studies. Chalasani and colleagues8 found that 1.4% of patients (7 of 517 patients) who had BCs drawn had therapy changed as a result of the BC result, and that 20.6% of patients (7 of 34 patients) with positive BC results had their treatment changed as a result. Waterer and colleagues,26 in a retrospective chart review of patients with confirmed pneumococcal pneumonia, found these figures to be 1.7% (31 of 1,805 patients) and 41.9% (31 of 74 patients), respectively.

Cost of BCs
Regarding the cost savings attributed to BCs in our study, a change to a less expensive antibiotic was attributed to the BC in 27.9% of positive BC results (12 of 43 patients), although a change to a more expensive regimen occurred in 14% of cases in disagreement with BC results (6 of 43 patients). Our costs were significantly less than the cost reported by authors in the United States.8 16 Chalasani et al8 reported a cost in US dollars of $4,874.57 per BC result that led to a change in antibiotic treatment.

Most of the cost associated with treating CAP is incurred by inpatient treatment.28 As a result, empiric outpatient regimens with a high oral bioavailability could realize considerable savings compared to efforts to save costs by trying to identify organisms that may permit step-down to less expensive drugs,29 or to hospital admission until adequate coverage has been ensured.

Epidemiologic Value of BCs
Antimicrobial therapy directed at a specific organism is considered ideal because of reduced costs, reduction in adverse drug reactions, and antimicrobial selection pressure.5 Most clinicians can quote an incident where a positive BC result "saved the day" in a difficult clinical situation.17 18 At this stage, BC results do allow for the best epidemiologic data; in the majority of cases, a positive BC result is definite evidence that that pathogen is the cause of the pneumonia, providing a far greater degree of certainty than that provided by sputum cultures or serology. Trends in the causes of bacteremic pneumonia, such as the current shift to S aureus as the second most common cause in Alberta (unpublished observations from T. J. Marrie, MD, February 2002), as well as important trends in antimicrobial resistance patterns, would not have been recognized without BC results.

However, this study highlights the weakness of BCs from an epidemiologic viewpoint. From the point of view of disease surveillance, only 43 of 1,022 study inpatients (4.2%) had pathogens identified by BC results. Even before considering that some of these may be false-positive findings, these organisms are not likely to be representative of those infecting the total population. Chlamydia pneumoniae and Mycoplasma pneumoniae are both well-known causes of CAP requiring hospitalization6 and, like viral pathogens, neither is identified by BC results,30 indicating that BCs drawn for epidemiologic reasons will miss important causes of CAP. In the case of Legionella, in which an etiologic diagnosis is considered particularly helpful, the majority of cases will be missed on BC, and a rapidly performed urinary antigen test has shown better sensitivity and specificity.6 13

Even if BCs are used specifically to identify S pneumoniae (the most frequent yield of BCs in most studies), they are not a particularly sensitive tool. Porath and colleagues31 found that only 16.9% of cases of S pneumoniae in their study were identified by BC results, while 83% of cases were identified by serology. It could be argued that serologic testing, although usually too slow to direct clinical management, has a higher and more specific yield, and would be better suited for disease surveillance. However, the specificity of serology (particularly for identifying S pneumoniae) has yet to be validated,31 and the important issue of drug resistance would be neglected. Recently published Canadian and American guidelines for the management of CAP recommend against serology as part of the routine management of CAP.5 6 The fact that S pneumoniae bacteremia is significantly less likely to be caused by penicillin-resistant strains suggests that the use of BCs to track trends in PRSP would tend to underestimate its incidence.15

Between 9.5% and 48% of pneumonias may be polymicrobial.25 26 30 32 33 34 35 36 37 38 Porath and colleagues31 found that 38.4% of patients (133 of 346 patients) with CAP had at least one other etiologic cause of CAP. Mundy and colleagues37 reported that in patients with confirmed atypical pathogens, 55.2% (16 of 29 patients) had concurrent infections diagnosed. As confirmed in this study, the highest yield of BC in CAP is most often of S pneumoniae. Of the cases of S pneumoniae identified by Porath et al,31 67.6% (100 of 148 patients) had co-pathogens identified, usually "atypicals." Although the clinical importance of polymicrobial infection is uncertain,5 37 some authors have found that mixed infections are associated with a more complicated course.39 Other studies40 41 have shown a less complicated course in patients in whom atypical organisms were covered by their empiric regimens, and Waterer et al42 found that therapy with a single effective antimicrobial agent in confirmed cases of bacteremic pneumococcal pneumonia resulted in a greater risk of death than therapy with two agents, even after excluding cases in which the organism cultured was resistant to one of the initial agents. At this stage, despite the fact that S pneumoniae is more likely to cause severe disease than atypical organisms,43 it is unclear as to whether the identification of S pneumoniae on BC, and the "narrowing" of spectrum of antimicrobial coverage to cover only that organism will allow for the most positive clinical outcome, given the possibility of untreated "atypical" co-pathogens,5 the possibility of improvement in antibacterial effectiveness as a result of synergistic interaction between agents, and the possibly beneficial nonantibacterial properties of certain antibiotics.42

Findings that in vitro susceptibility results do not directly correlate with the predicted in vivo efficacy of antibiotics44 45 46 draw into question the clinical utility of susceptibility testing on the few positive BC (or sputum) results that are obtained. The identification of in vitro resistance may in fact result in more expensive and broader-spectrum antibiotics being chosen in cases where the initial choice may well have been adequate. Current laboratory definitions of "high level" penicillin-resistant pneumococcus are those with minimum inhibitory concentration values > 2.0 µg/mL, although the breakpoint of clinically relevant resistance to penicillin has been identified as a minimum inhibitory concentration value > 4.0 µg/mL and, at this stage, high-dose ß-lactam therapy is still considered the drug of choice for PRSP in CAP in the absence of meningitis.43 Similarly, macrolide resistance in North America (in most cases due to an efflux mechanism) does, for the most part, seem to be more of a laboratory than a clinical phenomenon.47 This suggests that, although broadening the coverage of therapy as a reaction to BC sensitivities may be safer, it may also be unnecessary. Evidence in severe CAP suggests that delayed identification of an "incorrect" empiric regimen does not improve outcome.15 16 34 44 48

Relationship Between the Severity of CAP and BC Results
In our study, the severity of illness, as measured by the PSI, correlated poorly with the yield of BC. Admitted patients were equally likely to have positive BC results in classes I and II (in which mortality, as described by Fine et al,20 would be expected to be < 0.6%) as were patients in class V, with an expected mortality of 27% [7.2% and 6.9%, respectively]). This finding is in contrast to that found by other investigators.7 13 14 Waterer and Wunderink13 found that patients with a PSI class of IV or V are more likely to have a change in management directed by BC results (compared to no change in classes I to III).13 Theerthakarai and colleagues14 suggested that BC is without value in patients not meeting the 1993 American Thoracic Society guidelines5 criteria for severe CAP, while other authors have suggested that BC is useful as a prognostic tool in patients in whom mortality is likely.44 Bohte and colleagues35 found a relative risk of death from CAP to be 3.4 times higher in those with positive BC results compared with those with negative BC results.

Limitations
Firstly, as a sector of the population at risk for CAP was excluded, our findings should not be applied to patients with any of the conditions listed as exclusion criteria in the introduction. Similarly, as adult and pediatric pneumonias are etiologically distinct entities, with distinct management strategies,49 our findings should not be used to influence inpatient management of CAP in children.

Secondly, there was a difference in the percentage of BCs ordered in the two study arms. The clinical guideline used in the intervention arm included instructions to draw BCs on all patients. Almost double the proportion of patients in the intervention arm had BCs done, compared to the conventional arm (58% vs 33%, respectively). This difference likely reflects physicians changing their practice as a result of being instructed to by the study protocol. This is further supported by the fact that the baseline characteristics of the patients in the two study arms were found to be similar,19 as were the percentage yield of BCs in admitted patients in each study arm (5.53% for the intervention arm vs 5.81% for the conventional arm). The percentage of BC ordered in the conventional arm likely reflects "real-life" physician practice, and possibly indicates physician cynicism with respect to universal BC orders for admitted patients with CAP. Because the proportion of patients in our study having BCs drawn is likely more than in the "real world," the cost savings that can be estimated from this study may be exaggerated by including the cost of negative BC results that may not have been ordered under "nonstudy" circumstances. However, looking at the results in the conventional arm alone (the study arm in which no change in practice was prescribed), 40% of patients with positive BC results (8 of 20 patients) had changes in management made that were attributed to the BC result. This represents 2.3% of the total number of BCs drawn in that group, with a cost of $1,769.88 per case in which treatment was changed (compared to 1.7% of the total, and a cost of $2,446.08 per similar case in the intervention arm). This trend toward decreased clinical cost-effectiveness of BC as a greater proportion are done may indicate that physician judgment as to which patients might be most likely to benefit from BCs (the factor most likely to result in guidelines being followed in the real world) is accurate.

Regarding the issue of whether decisions to step down to less expensive drugs may have been influenced in the intervention arm by a study drug being prescribed by the pathway, the percentage of step-downs was found to be similar in both groups. In the conventional arm alone, failure to step down treatment appropriately in > 60% of patients demonstrates that physicians are not often guided by positive BC results that would decrease the cost of therapy, and that clinical observations are much more likely to lead to therapeutic decisions. This study may actually overestimate the value of BC, in that step-down therapy may well have been carried out on clinical grounds, even in the absence of BC results.

Finally, the intent of this study was not to perform a formal economic evaluation, and our crude estimate of costs presumes that the adjusted regimen would have followed the same sequence of IV and oral preparations. This does not account for the fact that certain antibiotics, because of more reliable oral bioavailability, may have been stepped down to oral routes of administration earlier than their substitutes.

The cost and adverse clinical outcome that could theoretically have been incurred by missing an organism not covered by the empiric regimen cannot be estimated from this study. However, numerous authors7 9 12 48 50 51 have shown that the achievement of a microbiological diagnosis does not improve patient outcome. Similarly, delayed diagnosis of an organism resistant to the empiric regimen has also not been found to improve outcome.34 42 Some authors have even reported an increase in costs as a result of false-positive BC findings.7 10


   Conclusion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
Appendix
 References
 
We conclude that our study adds to the argument that routine BCs rarely contribute significantly to the clinical management of CAP. We agree with other investigators7 11 that, in spite of guidelines suggesting routine BCs in hospitalized patients with CAP,1 2 3 4 5 6 and the fact that a positive BC result affords a degree of reassurance to clinicians, there is very little evidence that this reassurance is warranted or cost-effective. BCs should not form part of the routine management of CAP. We recognize, however, that at this stage and until more efficient and cost-effective diagnostic aids are developed, it may be prudent to continue to draw BCs in severe cases of CAP or in cases of CAP with any of the exclusion criteria listed. Outpatient management of CAP, using newer antibiotics with a high oral bioavailability and excellent coverage of the majority of CAP pathogens, may allow realization of significant savings compared to hospital admission pending the results of microbiological investigation of patients with CAP.


   Appendix
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
 Appendix
References
 
The capitaL Study investigators include the following: D. Gregson, J. Gibson, K. Woolfrey, C. Hammerberg (St. Joseph’s Health Centre, London, ON); A. McIvor, P. Hawkins, K. Johnson, K. Robinson (Sunnybrook Health Science Centre, Toronto, ON); R. Saginur, J.R. Worthington, K. Heney, G. Nichol (Ottawa Civic Hospital, Ottawa, ON); B. Feagan, E. Ralph, K. Theakston, C. McCabe (London Health Sciences Centre, University Campus, London, ON); G. Stiver, T. Lee, J. Buchanan, K. Sleigh, D. McKnight, A. Grunfeld, B. Kassen (Vancouver Hospital and Health Sciences Centre, Vancouver, BC); S. Affias, A. Coakly, B. MacDonald, S. Butler (Cape Breton Regional Hospital, Sydney, NS); M. Cook, D. Rowe, C. Bartlett, M. Howlett (Colchester Regional Hospital, Truro, NS); L. Harrigan, R. Dourcet, K Roddgues (Valley Regional Hospital, Kentville, NS); A. Dhar, A. McGee, D. Martin (Thunder Bay Regional Hospital, Port Arthur Site, Thunder Bay, ON); W. Patrick, S. Campbell, D. Sinclair, C. Touchie, K. Forward, G. Patrick, D. Murray, T. Romard (Queen Elizabeth II Health Sciences Centre, Halifax, NS); S. Houston, B. Rowe, A. Lindemulder, S. Roberts (University of Alberta Hospital, Edmonton, AL); G. Victor, G. Dickinson, A. Cwinn, L. Radey, D. Rehel-Giordano, K. Fyke, D. Garber, I. Sèquin (Ottawa General Hospital, Ottawa, ON); J. Hutchinson, V. Patel, J. Watson, E. Condon (Health Sciences Centre, St. John’s, Newfoundland); F. Smaill, L. Kelleher, S. Brons, S. Savoy (McMaster University Medical Centre, Hamilton, ON); G. Dow, C. MacLoughlin, G. Pavlatos-Jones, M. Hanley, J. Thompson (The Moncton Hospital, Moncton, NB); G. Sullivan, P. Legere, M. Wheeler (Western Regional Health Centre, Yarmouth, NS); A. Atkinson, L. Farrell, S. Atkinson (Health Services Association of the South Shore, Bridgewater, NS); P. Whitsitt, R. Vandersluis, D. Taylor (Oshawa General Hospital, Oshawa, ON); A. Holmes, C. Metcalfe (Royal Columbian Hospital, New Westminster, BC); L. Rance, R. Arshinoff, T. Smith, R. Anstey, U. Kuprath, J. Sigindere, J. Perkins, V. Holloway (Janssen-Ortho Inc., Toronto, ON); and E. Liddiard, W. Johnson (London Clinical Trials Research Group, London, ON).

This study was carried out at the following centers: St. Joseph’s Health Centre, London, ON; Sunnybrook Health Science Centre, Toronto, ON; Ottawa Civic Hospital, Ottawa, ON; Vancouver Hospital and Health Sciences Centre, Vancouver, BC; Cape Breton Regional Hospital, Sydney, NS; Colchester Regional Hospital, Truro, NS; Valley Regional Hospital, Kentville, NS; Thunder Bay Regional Hospital, Port Arthur Site, Thunder Bay, ON; QEII Health Sciences Centre, Halifax, NS; University of Alberta Hospital, Edmonton, AL; Ottawa General Hospital, Ottawa, ON; Health Sciences Centre, St. John’s, Newfoundland; McMaster University Medical Centre, Hamilton, ON; Moncton Hospital, Moncton, NB; Western Regional Health Centre, Yarmouth, NS; Health Services Association of the South Shore, Bridgewater, NS; Oshawa General Hospital, Oshawa, ON; Royal Columbian Hospital, New Westminster, BC; Janssen-Ortho, Toronto, ON; and London Clinical Trials Research Group, London, ON.


   Acknowledgements
 
We thank Cynthia Lank of Cynthia N. Lank Editorial Services and Heather McQuinn of the Queen Elizabeth II Health Sciences Centre Emergency Department for their help in preparing this manuscript.


   Footnotes
 
Abbreviations: BC = blood culture; CAP = community-acquired pneumonia; MRSA = methicillin-resistant Staphylococcus aureus; PRSP = penicillin-resistant Streptococcus pneumoniae; PSI = pneumonia severity index; PSSP = penicillin-susceptible Streptococcus pneumoniae

This study was supported by a joint Medical Research Council of Canada/industry grant (the industry partner was Jansen Ortho, manufacturer of levofloxacin), and by grant 9807PT-39621-UI-D from the Medical Research Council of Canada, Pharmaceutical Manufacturers’ Association of Canada Awards Program.

At the time of the capitaL Study, Rosemary Anstey was an employee of Janssen-Ortho. She is a shareholder of Johnson and

Johnson stock (Janssen-Ortho is a member company of Johnson and Johnson). Dr. Marrie currently has research grants from Bayer, Pfizer, and Abbott.

Received for publication May 10, 2002. Accepted for publication October 22, 2002.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
 Appendix
 References
 

  1. Bartlett, JG, Breiman, RF, Mandell, LA, et al (1998) Community-acquired pneumonia in adults: guidelines for management. Clin Infect Dis 26,811-838[ISI][Medline]
  2. Feldman, C, Bateman, ED, Klugman, KP, et al Management of community-acquired pneumonia in adults. S Afr Med J 1996;86,1152-1163[Medline]
  3. Finch, RG, Woodhead, MA Practical considerations and guidelines for the management of community-acquired pneumonia. Drugs 1998;55,31-45[CrossRef][ISI][Medline]
  4. San Pedro, GS, Campbell, GD Limitations of diagnostic testing in the initial management of patients with community-acquired pneumonia. Semin Respir Infect 1997;12,300-307[Medline]
  5. Niederman, MS, Mandell, LA, Anzueto, A, et al Guidelines for the management of community-acquired pneumonia: diagnosis, assessment of severity, antimicrobial therapy, and prevention. Am J Respir Crit Care Med 2001;163,1730-1754[Free Full Text]
  6. Mandell, LA, Marrie, TJ, Grossman, RF, et al Canadian guidelines for the initial management of community-acquired pneumonia: an evidence-based update by the Canadian Infectious Diseases Society and the Canadian Thoracic Society. Clin Infect Dis 2000;31,383-421[CrossRef][Medline]
  7. Levy, M, Dromer, F, Brion, N, et al Community-acquired pneumonia: importance of initial noninvasive bacteriologic and radiographic investigations. Chest 1988;92,43-48
  8. Chalasani, NP, Valdecanas, MAL, Gopal, AK, et al Clinical utility of blood cultures in adult patients with community-acquired pneumonia without defined underlying risks. Chest 1995;108,932-938[Abstract/Free Full Text]
  9. Hickey, RW, Bowman, MJ, Smith, GA Utility of blood cultures in pediatric patients found to have pneumonia in the emergency department. Ann Emerg Med 1996;27,721-725[CrossRef][ISI][Medline]
  10. Bates, DW, Goldman, L, Lee, TH Contaminant blood cultures and resource utilization: the true consequences of false-positive results. JAMA 1991;25,365-369
  11. Woodhead, MA, Arrowsmith, J, Chamberlain-Webber, R, et al The value of routine microbiological investigation in community-acquired pneumonia. Respir Med 1991;85,313-317[ISI][Medline]
  12. Ewig, S, Bauer, T, Hasper, E, et al Value of routine microbial investigation in community-acquired pneumonia treated in a tertiary care centre. Respiration 1996;63,164-169[ISI][Medline]
  13. Waterer, GW, Wunderink, RG The influence of the severity of community-acquired pneumonia on the usefulness of blood cultures. Respir Med 2001;95,78-82[CrossRef][ISI][Medline]
  14. Theerthakarai, R, El-Halees, W, Ismail, M, et al Non-value of the initial microbiological studies in the management of non-severe community-acquired pneumonia. Chest 2001;119,181-184[Abstract/Free Full Text]
  15. Wunderink, RG, Waterer, GW Appropriate microbiological testing in community-acquired pneumonia. Chest 2001;119,5-7[Free Full Text]
  16. Sanyal, S, Smith, PR, Saha, AC, et al Initial microbiologic studies did not affect outcome in adults hospitalized with community-acquired pneumonia. Am J Respir Crit Care Med 1999;160,346-348[Abstract/Free Full Text]
  17. Berk, SL Justifying the use of blood cultures when diagnosing community-acquired pneumonia. Chest 1995;108,891-892[Free Full Text]
  18. Bryan, CS Blood cultures: no place to skimp. Chest 1999;116,1153-1154[Free Full Text]
  19. Marrie, TJ, Lau, CY, Wheeler, SL, et al A controlled trial of a critical pathway for treatment of community-acquired pneumonia. JAMA 2000;283,749-755[Abstract/Free Full Text]
  20. Fine, MJ, Auble, TE, Yealy, DM, et al A prediction rule to identify low-risk patients with community-acquired pneumonia. N Engl J Med 1997;336,243-250[Abstract/Free Full Text]
  21. Nova Scotia laboratory credit schedule, April 1, 1994. Ref: #940510. 1994 Nova Scotia Department of Health and Fitness. Halifax, Nova Scotia:
  22. Lynch, JP, Martinez, FN Community-acquired pneumonia. Curr Opin Pulm Med 1998;4,162-172[Medline]
  23. Marrie, TJ Community-acquired pneumonia. Clin Infect Dis 1994;18,501-515[ISI][Medline]
  24. Marrie, TJ, Fine, MJ, Obrosky, DS, et al Community-acquired pneumonia due to Escherichia coli. Clin Microbiol Infect 1998;4,717-723[Medline]
  25. Marston, BJ, Plouffe, JF, File, TM, Jr, et al Incidence of community-acquired pneumonia requiring hospitalization: results of a population-based active surveillance study in Ohio. Arch Intern Med 1997;157,1709-1718[Abstract]
  26. Waterer, GW, Jennings, SG, Wunderink, RG The impact of blood cultures on antibiotic therapy in pneumococcal pneumonia. Chest 1999;116,1278-1281[Abstract/Free Full Text]
  27. Glerant, JC, Hellmuth, D, Schmit, JL, et al Utility of blood cultures in community-acquired pneumonia requiring hospitalization: influence of antibiotic treatment before admission. Respir Med 1999;93,208-212[CrossRef][ISI][Medline]
  28. Niederman, MS, McCombs, JS, Unger, AN, et al The cost of community-acquired pneumonia. Clin Ther 1998;20,820-837[CrossRef][ISI][Medline]
  29. Ramirez, JA Switch therapy with ß-lactam/ß-lactamase inhibitors in patients with community-acquired pneumonia. Ann Pharmacother 1998;32(suppl),S22-S26[Abstract/Free Full Text]
  30. Bates, JH, Campbell, GD, Barron, AL, et al Microbial etiology of acute pneumonia in hospitalized patients. Chest 1992;101,1005-1012[Abstract/Free Full Text]
  31. Porath, A, Schlaeffer, F, Pick, N, et al Pneumococcal community-acquired pneumonia in 148 hospitalized adult patients. Eur J Clin Microbiol Infect Dis 1997;16,863-870[CrossRef][ISI][Medline]
  32. Mundy, LM, Auerwater, PG, Oldach, D, et al Community-acquired pneumonia: impact of immune status. Am J Respir Crit Care Med 1995;152,1309-1315[Abstract]
  33. Lieberman, D, Schlaeffer, F, Boldhur, I, et al Multiple pathogens in adult patients admitted with communityacquired pneumonia: a one-year prospective study of 346 consecutive patients. Thorax 1996;51,179-184[Abstract]
  34. Ewig, S, Ruiz, M, Torres, A, et al Pneumonia acquired in the community through drug-resistant Streptococcus pneumoniae. Am J Respir Crit Care Med 1999;159,1835-1842[Abstract/Free Full Text]
  35. Bohte, R, van Furth, R, van den Broek, PJ Aetiology of community-acquired pneumonia: a prospective study among adults requiring admission to hospital. Thorax 1995;50,543-547[Abstract]
  36. Wunderink, RG Diagnosis of pneumonia. Curr Opin Pulm Med 1996;2,213-217[Medline]
  37. Mundy, LM, Oldach, D, Auerwarter, PG, et al Implications for macrolide treatment in community-acquired pneumonia. Chest 1998;113,1201-1206[Abstract/Free Full Text]
  38. Ruiz, M, Ewig, S, Marcos, MA, et al Etiology of community-acquired pneumonia: impact of age, comorbidity, and severity. Am J Respir Crit Care Med 1999;160,397-405[Abstract/Free Full Text]
  39. Kauppinen, M, Saikku, P, Kujala, P, et al Clinical picture of community-acquired Chlamydia pneumoniae requiring hospital treatment: a comparison between chlamydial and pneumococcal pneumonia. Thorax 1996;51,185-189[Abstract]
  40. Gleason, PP, Meehan, TP, Fine, JM, et al Associations between initial antimicrobial therapy and medical outcomes for hospitalized elderly patients with pneumonia. Arch Intern Med 1999;159,2562-2572[Abstract/Free Full Text]
  41. Gleason, PP, Kapoor, WN, Stone, RA, et al Medical outcomes and antimicrobial costs with the use of the American Thoracic Society guidelines for outpatients with community-acquired pneumonia. JAMA 1997;278,32-39[Abstract]
  42. Waterer, GW, Somes, GW, Wunderink, RG Monotherapy may be suboptimal for severe bacteremic pneumococcal pneumonia. Arch Intern Med 2001;161,1837-1842[Abstract/Free Full Text]
  43. Heffelfinger, JD, Dowell, SF, Jorgensen, JH, et al Management of community-acquired pneumonia in the era of pneumococcal resistance. Arch Intern Med 2000;160,1399-1408[Abstract/Free Full Text]
  44. Leroy, O, Santre, C, Beuscart, C, et al A five-year study of severe community-acquired pneumonia with emphasis on prognosis in patients admitted to an intensive care unit. Intensive Care Med 1995;21,24-31[CrossRef][ISI][Medline]
  45. Craig, WA Pharmokinetic/pharmacodynamic parameters: rationale for antibacterial dosing of mice and men. Clin Infect Dis 1998;26,1-10[ISI][Medline]
  46. Siegel, RE The significance of serum vs tissue levels of antibiotics in the treatment of penicillin-resistant and community-acquired pneumonia: are we looking in the wrong place? Chest 1999;116,535-538[Free Full Text]
  47. Amsden, GW Pneumococcal macrolide resistance: myth or reality? J Antimicrob Chemother 1999;44,1-6[Abstract]
  48. Feldman, C, Ross, S, Mahomed, AG, et al The aetiology of severe community-acquired pneumonia and its impact on initial, empiric, antimicrobial chemotherapy. Respir Med 1995;89,187-192[CrossRef][ISI][Medline]
  49. Metlay, JP, Hofmann, J, Cetron, MS, et al Impact of penicillin susceptibility on medical outcomes for adult patients with bacteremic pneumococcal pneumonia. Clin Infect Dis 2000;30,520-528[CrossRef][ISI][Medline]
  50. Rubins, JB, Janoff, EN Community-acquired pneumonia: tailoring management of adult patients according to risk category. Postgrad Med 1997;102,45-62
  51. Moine, P, Vercken, JB, Chevret, S, et al Severe community-acquired pneumonia: etiology, epidemiology, and prognosis factors. Chest 1994;105,1487-1495[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 ChestHome page
J. M. Porcel, A. Ruiz-Gonzalez, M. Falguera, A. Nogues, C. Galindo, J. Carratala, and A. Esquerda
Contribution of a Pleural Antigen Assay (Binax NOW) to the Diagnosis of Pneumococcal Pneumonia
Chest, May 1, 2007; 131(5): 1442 - 1447.
[Abstract] [Full Text] [PDF]

Home page
 JWatch Emergency Med.Home page
The CMS Blood Cultures for CAP Program: The Architects Speak Out
Journal Watch Emergency Medicine, April 27, 2005; 2005(427): 1 - 1.
[Full Text]

Home page
 PediatricsHome page
K. D. Schultz, L. L. Fan, J. Pinsky, L. Ochoa, E. O. Smith, S. L. Kaplan, and M. L. Brandt
The Changing Face of Pleural Empyemas in Children: Epidemiology and Management
Pediatrics, June 1, 2004; 113(6): 1735 - 1740.
[Abstract] [Full Text] [PDF]

Home page
 ThoraxHome page
J T Macfarlane and D Boldy
2004 update of BTS pneumonia guidelines: what's new?
Thorax, May 1, 2004; 59(5): 364 - 366.
[Full Text] [PDF]

Home page
 Am. J. Respir. Crit. Care Med.Home page
D. E. Craven
Blood Cultures for Community-Acquired Pneumonia: Piecing Together a Mosaic for Doing Less
Am. J. Respir. Crit. Care Med., February 1, 2004; 169(3): 327 - 328.
[Full Text] [PDF]

Home page
 Am. J. Respir. Crit. Care Med.Home page
M. L. Metersky, A. Ma, D. W. Bratzler, and P. M. Houck
Predicting Bacteremia in Patients with Community-Acquired Pneumonia
Am. J. Respir. Crit. Care Med., February 1, 2004; 169(3): 342 - 347.
[Abstract] [Full Text] [PDF]

Home page
 JWatch Emergency Med.Home page
More Debate on the Utility of Blood Cultures in CAP
Journal Watch Emergency Medicine, August 27, 2003; 2003(827): 2 - 2.
[Full Text]

Home page
 JWatch Infect. DiseasesHome page
Blood Cultures: Helpful in Managing Community-Acquired Pneumonia?
Journal Watch Infectious Diseases, June 20, 2003; 2003(620): 7 - 7.
[Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (43)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Campbell, S. G.