NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Soni NJ, Samson DJ, Galaydick JL, et al. Procalcitonin-Guided Antibiotic Therapy [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2012 Oct. (Comparative Effectiveness Reviews, No. 78.)

Cover of Procalcitonin-Guided Antibiotic Therapy

Procalcitonin-Guided Antibiotic Therapy [Internet].

Show details

Summary and Discussion

Summary of the Main Findings

There were 18 randomized controlled trials (RCTs) that compared procalcitonin guidance with the use of clinical criteria to manage antibiotic therapy in patients with known or suspected infection, or at risk of infection. The evidence addressed five patient populations: (1) seriously ill adult patients in the intensive care unit (ICU), including patients with ventilator-associated pneumonia (VAP) and those critically ill with suspected bacterial infections, severe sepsis, and septic shock; (2) patients with symptoms and signs of various respiratory tract infections; (3) neonates with suspected sepsis; (4) children between 1 to 36 months of age with fever of unknown etiology; and (5) postoperative patients without clinical evidence of infection.

Intensive Care Unit Patients

Five trials32-35,39 (n=938) addressed procalcitonin-guided discontinuation of antibiotic therapy. Strength of evidence was judged to be high that procalcitonin guidance reduces antibiotic use. The absolute difference in duration of antibiotic use, the measure that was reported by all five studies,32-35,39 ranged from -1.7 to -5 days, with relative reductions that ranged from 21 to 38 percent. There is moderate evidence that procalcitonin-guided antibiotic discontinuation does not increase morbidity as indicated by ICU and hospital length of stay. The strength of evidence was judged low that procalcitonin guided antibiotic therapy in the ICU does not increase mortality. Evidence on mortality was initially rated as stronger but was downgraded to low based on uncertainty about the appropriate noninferiority margin for this outcome. Although a 10 percent noninferiority margin has been recommended by the Infectious Diseases Society of America (IDSA) and American College of Chest Physicians (ACCP) in relevant populations, there is also strong sentiment that this is too high a noninferiority margin for mortality, which was expressed among our peer reviewers and also in the published literature. Moreover, there are presently two large trials in progress, which may in the future yield more precise estimates of mortality.

There is moderate evidence that procalcitonin-guided intensification of antibiotic therapy that broadens the spectrum of bacterial coverage does not improve outcomes in critically ill patients, and in fact, may have adverse consequences. The large (n=1,200), high-quality trial by Jensen et al.28 found greater duration and increased total exposure to antibiotics with procalcitonin guidance. There was also increased morbidity, including a 1 day increase (p=0.004) in ICU length of stay (LOS), a significant increase in days on mechanical ventilation, and increased days with abnormal renal function. A second fair quality study45 (n=72) was judged to be too small to be informative.

Reduced antibiotic use should decrease the number of allergic reactions, antibiotic-related side effects and toxicities, and superinfections, including the emergence of multi-drug resistant (MDR) organism and Clostridium difficile. Only one study39 reported on the emergence of MDR organisms and superinfections.

Respiratory Tract Infection

Eight trials27,31,36-38,40-42 (n=3,492) addressed initiation and/or discontinuation of antibiotics in patients with acute upper and lower respiratory tract infection. Settings included primary care clinics, emergency department (ED), and hospital wards. There is high strength of evidence that procalcitonin guidance reduces antibiotic duration and prescription rates; and moderate evidence of reduction in total antibiotic exposure. Absolute reduction in duration of antibiotic therapy ranged from 1 to 7 days with relative reductions ranging from -13 to -55 percent. Absolute reduction in prescription rates ranged from -2 to -7 percent with relative reductions ranging from -1.8 to -72 percent. There was moderate evidence that procalcitonin guidance did not increase mortality, hospital length of stay, or ICU admission rates. However, a limitation of the evidence is the very large number of study participants that would be required to narrow the confidence interval for estimated mortality. There was insufficient evidence to judge effects on days of restricted activity or on antibiotic adverse events. Three studies27,40,42 reported on adverse antibiotic effects, and there was a statistically significant reduction in the procalcitonin-guided arm versus the control arm that was associated with reduced antibiotic use. There was no consistency, however, on how adverse effects were defined and details on the types of adverse reactions were lacking. Only one study42 reported that the reduction in adverse antibiotic adverse effects was mainly due to a reduction in diarrhea. A more uniform approach to evaluating and reporting adverse effects related to antibiotic use would be useful in future studies.

Neonatal Sepsis

One good quality43 study (n=121) provided moderate evidence that procalcitonin guidance reduces the use of antibiotic therapy for suspected early neonatal sepsis. The duration of antibiotic use was overall reduced by 22.4 hours (22.0%). Further, the proportion of neonates on antibiotics for longer than 72 hours was reduced by 27 percent. Greatest reductions were seen among neonates who were judged according to clinical criteria to have possible infection or unlikely to have infection as compared with those with proven or probable infection. Strength of evidence was judged insufficient to make conclusions on mortality and morbidity due to small study size.

Fever of Unknown Source in Children Ages 1–36 Months

The strength of evidence was judged insufficient to draw conclusions on outcomes of procalcitonin-guided antibiotic therapy for fever of unknown source in children 1–36 months of age. One good quality RCT26 (n=384) reported no significant results.

Postoperative Patients at Risk of Infection

The strength of evidence was judged insufficient to draw conclusions on outcomes of procalcitonin-guidance to determine preemptive antibiotic therapy for patients after colorectal surgery. The evidence consisted of one small (n=20) trial.44

Clinical Context and Applicability of Evidence for Decisionmaking

The diagnosis of sepsis is challenging because the clinical criteria for the diagnosis overlap with noninfectious causes of systemic inflammation, such as the systemic inflammatory response syndrome. Initiation of antibiotic therapy for sepsis is necessary even while the diagnostic evaluation is ongoing since delayed antibiotic therapy is associated with increased mortality.14,48,50 A biomarker, such as procalcitonin, that improves decisions about initiating, discontinuing, or changing antibiotic therapy, could have substantial clinical benefits. This systematic review found that procalcitonin guidance reduces antibiotic use for adult patients in both medical and surgical ICUs. Generalizing results from the studies reviewed here, conducted primarily in Europe, would depend on similar use of and adherence with trial-based algorithms guiding antibiotic decision-making. Studies included patients who had co-morbidities that are common in the ICU patients (e.g., cardiac disease, diabetes, chronic lung disease, cirrhosis, chronic renal failure, cancer), and thus, the evidence from these studies are applicable to clinical practice in the ICU setting.

Respiratory tract infections contribute significantly to the problem of antibiotic misuse. Approximately 75 percent of all antibiotics prescribed in the ambulatory setting are for acute respiratory tract infections, but the vast majority of these infections are viral and do not benefit from antibiotic treatment.51 Clinical and microbiological evaluations are neither sensitive nor specific to differentiate bacterial from viral respiratory tract infections. Because procalcitonin levels rise soon after the onset of a bacterial infection, procalcitonin can help to differentiate bacterial from viral infections. Our systematic review found that procalcitonin guidance for initiation and discontinuation of antibiotic therapy significantly reduced antibiotic prescription rates and duration in patients with acute respiratory tract infections, including acute exacerbations of chronic obstructive pulmonary disease, community acquired pneumonia (CAP) and acute bronchitis. In most of these studies, antibiotic therapy was encouraged if the procalcitonin level was greater than 0.25 ng/mL, because a bacterial infection was likely. Similarly antibiotic therapy was discouraged if the procalcitonin level was less than 0.25 ng/mL because a bacterial infection was unlikely.

Certain populations, however, were excluded from one or more studies of procalcitonin guidance reviewed in this report. These groups might be considered high risk for increased morbidity and/or mortality with delayed initiation or shorter courses of antibiotic therapy, or may not have the same procalcitonin rise in response to infection due to their co-morbidities. Thus, findings from this review should not be extrapolated to these high-risk groups, which include pregnant patients; patients with absolute neutropenia; and other immunocompromised populations (solid organ and stem-cell transplant recipients, patients with advanced HIV infection/AIDS). Although such patients were excluded in these studies, future studies may help to determine if procalcitonin-guided antibiotic therapy is beneficial in these groups, as well. Although febrile neutropenic patients are usually continued on antibiotics until the neutropenia resolves, the most recent guidelines suggest that patients can be switched to an oral fluoroquinolone when an infection has been adequately treated, and procalcitonin guidance could potentially be used in this context.52 Of note, patients with chronic infections and infections where a longer duration of antibiotic therapy is standard of care, such as infective endocarditis, were also appropriately excluded from these studies.

Applicability to pediatric settings is a significant gap in the present evidence. Only two RCTs26,43 reported on procalcitonin guidance in pediatric populations. One study43 included neonates with suspected early sepsis. While antibiotic use was reduced, the trial was underpowered for morbidity and mortality outcomes. In fact, there were no mortality events in either arm of the study and only 21 of 121 neonates in the study had a probable or proven infection.43 The second study26 evaluated procalcitonin-guided antibiotic therapy in children ages 1–36 months presenting to the ED with fever of unknown etiology. No significant differences were observed for measures of antibiotic use, morbidity, or mortality with procalcitonin guidance. The evidence from this single study was judged insufficient to reach conclusions about the use of procalcitonin guidance in this setting. It is important to note that bacterial infections were uncommon in the study population. Only 25 percent of children were admitted to the hospital and about 27 percent received antibiotics overall. This rate was even lower, 10 and 9 percent respectively, if the children with neutropenia or a serious bacterial infection were excluded. There were no studies in children ages 3 to 18 years.

Ultimately, the value of procalcitonin-guided antibiotic therapy depends on the clinical benefits of reduced antibiotic use, which is difficult to quantify. Immediate consequences may include decrease in allergic reactions, drug toxicities, and frequency of C. difficile infection. A major downstream effect of reducing antibiotic use may be a lower probability of emergence of antibiotic-resistant strains. Antimicrobial resistance contributes to morbidity, mortality, and health care costs. Though infection control programs reduce transmission of resistant bacteria between patients, they do not affect development of resistance which occurs, in part, due to antibiotic overuse. There is some evidence that the development of resistance is more related to antibiotic treatment than transmission from patient to patient.53,54 There are several studies and indirect lines of evidence that suggest that control of antibiotic use can reduce emergence of resistance, but the data are limited.46 Reductions in antibiotic course duration have been associated with significant reductions in antibiotic adverse effects, C. difficile colitis, and superinfection with multidrug-resistant Gram-negative rods.46-48

In our systematic review, few studies reported on allergic and adverse events of antibiotic use 27,40,42 and only one32 reported on antibiotic resistance. The durability in reduction of antibiotic use is not addressed in these trials which limits their applicability to clinical practice. The setting of a clinical trial, or highly visible introduction of a new practice, can have a halo effect on physician behavior so the present evidence does not address long-term impact of using procalcitonin guidance a in real-world clinical setting. Antibiotic stewardship programs are now recommended for all institutions and there are guidelines for how they should function.55 Antibiotic stewardship programs are associated with reduced antibiotic use and also decreased adverse effects of antibiotic therapy. The evidence in this review does not compare outcomes of using procalcitonin guidance versus antibiotic stewardship programs nor does it address whether addition of procalcitonin to an antibiotic stewardship programs improves outcomes. There is at least one report56 that indicates the use of procalcitonin measurements may be used as part of an antibiotic stewardship program to decrease the duration of antibiotic therapy. Antibiotic stewardship activities are usually limited to the acute care hospital setting. Although it would be difficult or impractical for antibiotic stewardship programs to have active intervention in the outpatient setting, the use of procalcitonin might complement other types of outpatient programs, such as educational programs for physicians and patients aimed at reducing the use of antibiotics for respiratory tract infections.51

Discussion of Present Findings in Context of Other Systematic Reviews

We are aware of four systematic reviews3,19-21 that were published prior to our present review; the findings of our review are discussed in the context of these prior reviews. All of the previous reviews (including the present review) came to similar conclusions: procalcitonin-guided antibiotic decisionmaking compared with clinical criteria-guided antibiotic decision making reduces antibiotic use and is not associated with increased mortality or morbidity. We reviewed all published RCTs of the use of procalcitonin-guided initiation or discontinuation of antibiotic therapy, as well as studies that used procalcitonin for other interventions in patients with infection and/or sepsis. A total of 18 RCTs were included in our systematic review. As the most recent systematic review, ours is the only one that includes the large (n=1,200) high quality trial by Jensen et al.28 The Jensen trial28 is unique in showing that procalcitonin-guided intensification of antibiotic therapy to broaden the spectrum of bacterial coverage does not improve outcomes in critically ill patients, and in fact, may have adverse consequences. Among all the systematic reviews, only ours distinguished between procalcitonin-guided antibiotic intensification therapy, in contrast to procalcitonin-guided antibiotic initiation or discontinuation therapy.

As Table 31 shows, our systematic review differs from previous systematic reviews in terms of number of studies included, scope of indications addressed, and how populations were grouped for clinical relevance. The number of trials included in previous systematic reviews ranged from seven trials3,20,21 to 14 trials.19 While the three systematic reviews3,20,21 that each included seven studies overlapped to a large degree, none included the same studies. Agarwal and Schwartz included one published abstract57 not included in any other systematic review. We did not include the published abstract by Layios57 in our review, since it was not possible to assess study quality from an abstract. Tang3 included patients in ICUs and patients with respiratory tract infections, but did not analyze the data from each population separately. Kopterides21 focused on studies in the ICU population but their meta-analysis of antibiotic use outcomes pooled data from neonatal population with the adult population. A strength of the Scheutz review19 was separate meta-analyses of mortality based on acuity of illness (primary care, ED, and ICU patients), as well as a meta-analysis for the total population. The authors concluded that procalcitonin guidance reduces antibiotic duration in primary care, ED, and ICU patients, and reduces prescription rates only in lower acuity primary care and ED patients. The Scheutz review included the 2009 study by Long,58 whereas we had access to an overlapping, but more recent report of this study.31

Table 31. Summary of systematic reviews.

Table 31

Summary of systematic reviews.

A strength of the present review is that it addresses pediatric populations separately from adult patients. In addition, it also recognizes that there are distinct patient groups within the pediatric population stratified by age. In our review, we separately grouped neonates and children ages 1–36 months, each represented by a single study. There were no studies in children ages 3 years to 18 years.

Limitations of Present Review

A challenging aspect of this review was appraising the strength of evidence that procalcitonin-guided therapy did not result in any increased morbidity or mortality in the critically ill and respiratory tract infection populations. In the studies of critically ill patients where procalcitonin was used to reduce antibiotic exposure, only the Bouadma study39 did a power analysis and used a predefined a margin for noninferiority for 28- and 60-day mortality. Meta-analysis was performed looking at early mortality across all five ICU studies. Results show a pooled point estimate of 0.4 percentage-point reduction in mortality, and the 95% confidence interval (CI) for the difference in mortality between procalcitonin-guided therapy and standard care was between -6 percent and 5 percent favoring the procalcitonin-guided therapy group. There is disagreement, however, whether or not this falls within the appropriate noninferiority margin. The choice of a noninferiority margin only requires sufficient precision to exclude a minimal important difference (MID).49 Although a 10 percent noninferiority margin for mortality has been recommended by the IDSA and ACCP in relevant populations, there is concern, expressed by some peer reviewers and in literature, that a 10 percent margin may be too high. Initially, a higher strength of evidence was considered but due to this uncertainty of the noninferiority margin, the strength of evidence was downgraded to low that procalcitonin-guided antibiotic therapy in the ICU does not increase mortality. Even though overall strength of evidence was low, the results were judged to be precise because the pooled point estimate was centered on the null and the 95% CI was narrow (11 percentage points). While only one study was powered for mortality, one purpose for meta-analysis is to overcome insufficient power and the group of studies was highly consistent: statistical heterogeneity, as expressed by the I2 statistic, was found to be 0 percent. Sixty-day mortality was reported by one study39 and was not included in our analysis because late mortality is more likely related to underlying comorbidities. Moreover, there are presently two large trials in progress, which may in the future yield more precise estimates of mortality.

Our systematic review compared procalcitonin guidance with antibiotic therapy based on usual clinical criteria, algorithms, or guidelines. In view of the present emphasis on the overuse of antibiotics, other interventions to reduce antibiotic use, such as institution of antibiotic stewardship programs and implementation of practice guidelines in institutional settings, may have been more robust comparators by which to assess the outcomes of procalcitonin guidance. A limitation of our review is that we did not systematically seek evidence comparing procalcitonin guidance with antibiotic stewardship programs or other programs aimed at reducing antibiotic use. We also did not assess studies that have implemented procalcitonin-guided antibiotic therapy into an antibiotic stewardship program.

Future Research

Summary of Weaknesses or Gaps in the Evidence

This systematic review compared outcomes of procalcitonin guidance compared with clinical criteria alone to initiate, discontinue or intensify antibiotic therapy. We identified five gaps in the evidence related to specific populations or comparators. We also identified methodologic weaknesses that were common across the studies and bodies of evidence reviewed in this report.

Research Gap 1. What Are the Outcomes of Procalcitonin Guidance in Subgroups of Patients Who Are Immunocompromised?

Patients with certain conditions were excluded from these studies, including neutropenia and immunocompromised states (solid organ and stem-cell transplant recipients, and patients with advanced HIV infection). The reasons for excluding such patients were not specifically stated. Because procalcitonin levels are affected by the host cytokine response to infection, the procalcitonin cutoff levels are most likely to differ in these populations. These populations are often excluded from clinical trials because these groups may be at higher risk of adverse outcomes. Finally, for some groups, such as neutropenic patients, antibiotics are continued until the neutropenia resolves, rather than until there is a clinical resolution.

Immunocompromised patients often comprise a significant portion of the ICU population, and in the large PRORATA39 study, where immunocompromised patients were included, they accounted for 16.6 percent of the population. In the PROVAP35 study of ventilator-associated pneumonia, 27.9 percent of the eligible patients were excluded because of immunosuppression. Even in community respiratory tract infections, such as CAP (7.6% excluded) and even in other respiratory tract infections (2.5% excluded), there is a significant subpopulation of patients who are immunocompromised or have condition such as cystic fibrosis where the efficacy and safety of procalcitonin-guided therapy is unknown.36,37

Six31,36-38,40,42 of the eight27,31,36-38,40-42 studies evaluating procalcitonin guidance in patients with acute respiratory tract infections specifically excluded immunocompromised patients. While severely immunocompromised patients presenting with clinical signs of infection are most likely treated empirically with antibiotics, patients with mild to moderate immunosuppression, such as patients on low-dose corticosteroids for chronic inflammatory conditions, may not benefit from antibiotic therapy, even though they are often treated empirically. Procalcitonin guidance may have a potential role in reducing antibiotic use in the ambulatory patients with mild to moderate immunosuppression compared with standard therapy.

Exclusion of immunocompromised patients was common across all five patient populations in this review. Immunocompromised patients are considered high risk for increased morbidity and/or mortality with delayed initiation or shorter courses of antibiotic therapy, or may not have the same procalcitonin rise in response to infection due to their co-morbidities. However, immunocompromised patients may also gain significant clinical benefits if their antibiotic use can be safely reduced, because they are vulnerable to mortality and morbidity from antibiotic resistance and adverse effects of antibiotics.

In addition to immunocompromised patients, future research in needed in patients with certain local or systemic infections, such as skin and soft tissue infections and osteomyelitis. These are relatively common infections in which initiation and/or duration of antibiotic therapy is often unclear, and procalcitonin guidance may have a potential role in reducing antibiotic use.

Research Gap 2. What Are the Outcomes of Procalcitonin Guidance in Pediatric Patients?

Only two studies26,43reported on procalcitonin guidance in pediatric populations, and both were underpowered to assess morbidity and mortality outcomes. Both studies were limited to the acute care hospital setting. In our review, we grouped separately neonates43 and children26 ages 1-36 months, each represented by a single study. There were no studies in children ages 3 years to 18 years. Future studies of procalcitonin-guided initiation and discontinuation of antibiotics in the pediatric population will be extremely important. The overuse of antibiotics in pediatrics, in both the inpatient and outpatient setting, is as important among children, as it is in adults.

Research Gap 3. What Are the Outcomes of Procalcitonin Guidance in Identifying Patients at Risk of Infection Who Might Benefit From Preemptive Antibiotic Therapy?

The study by Chromik et al.44 reported that procalcitonin levels could accurately identify a subpopulation, 8 percent of patients who underwent elective colorectal surgery, who were at risk of a local or systemic infection. Although this was a small study, the evidence suggests that this approach might identify a group who would benefit from preemptive antibiotic therapy given before any infection is clinically evident. Larger studies are needed to confirm that preemptive antibiotic therapy can reduce infectious complications. Other patient populations who are at risk for infectious complications include burn patients, ICU patients, and postoperative patients who have undergone procedures other than colorectal surgery.

Research Gap 4. Does the Use of Procalcitonin Guidance Reduce Antibiotic Resistance and Antibiotic Adverse Events?

Although the importance of reducing antibiotic use is recognized and accepted, there was insufficient evidence from the RCTs we reviewed that the observed reduction in antibiotic use had any benefits with respect to antibiotic adverse reactions, superinfections, or the development of resistance. Adverse antibiotic effects were reported in only three studies27,40,42 and the findings were reported in different ways. Only one study39 reported on the emergence of multidrug resistant bacteria, with no differences found between procalcitonin-guided and standard antibiotic therapy, and none reported on the incidence of C. difficile. When designing future studies, there should be consideration for standardized reporting of adverse events from antibiotics, the incidence of C. difficile, and active surveillance for colonization with drug-resistant pathogens. Quality of life was not addressed in any of the randomized controlled studies included in this review, and if future research demonstrates a significant reduction in antibiotic adverse events, addressing quality of life would be an important consideration to include in future trials.

Research Gap 5. How Does Procalcitonin-Guided Antibiotic Therapy Compare With Other Approaches for Reducing Unnecessary Antibiotic Use, Such as Antibiotic Stewardship Programs and Implementation of Practice Guidelines?

In view of the present emphasis on the overuse of antibiotics, other interventions to reduce antibiotic use, such as institution of antibiotic stewardship programs and structured implementation of practice guidelines, may have been more robust comparators by which to assess the outcomes of procalcitonin-guided decisions on initiation and discontinuation of antibiotic therapy. Our review did not systematically seek evidence comparing procalcitonin guidance with other interventions to reduce antibiotic use, or evidence assessing whether addition of procalcitonin guidance to other interventions improves outcomes. So a systematic review that addresses a broader range of comparators is likely an initial step to determine whether there is an evidence gap and the nature of any gap or gaps. Given the urgency of reducing unnecessary use of antibiotics, there may be promising opportunities for future research that can inform clinical practice.

In addition to the research gaps above, we also identified four important methodologic weaknesses that were common across the studies and bodies of evidence reviewed in this report.

Summary of Methodological Weaknesses in the Evidence

Weakness 1. Measurement of Total Antibiotic Exposure

Total antibiotic exposure is used to capture the patient's total exposure to all antibiotics, and it is calculated by multiplying the total number of antibiotics by the number of days the patient is receiving each of the antibiotics. Total antibiotic exposure is conventionally reported as mean days per 1,000 days of followup, but some of the studies in this review only reported relative or absolute differences. The differences in reporting limited our ability to pool the total antibiotic exposure data in this review. Consistent use of the conventional measure of mean days per 1,000 days of followup would improve accumulation of a robust body of evidence on the outcomes of procalcitonin guidance.

Weakness 2. Measurement of Morbidity

There were various measures of morbidity across these studies, and that was true even between studies that were grouped together because of their similarities. Although admission rates, LOS, and ICU LOS were easy to compare, other measures were not. In the ICU populations, for example, the need for mechanical ventilation was often reported differently and studies used a variety of severity of illness scores (SOFA, SAP II, SAP III, and APACHE II). This makes it difficult to compare or pool data across studies.

Weakness 3. Rationale for Noninferiority Margins for Studies of Mortality

Mortality rates in trials of procalcitonin-guided therapy implicitly or explicitly pose a question of noninferiority, that is, can reduction in antibiotic use be achieved without a deleterious impact on survival? Is mortality no worse than with usual care? The choice of a noninferiority margin incorporates clinical and statistical judgments.59 The PRORATA39 trial, which was the largest trial (n=621) of procalcitonin used to discontinue antibiotic in the ICU populations, was designed to have an 80 percent power to exclude a 10 percent difference in mortality between groups. However, there is concern that the 10 percent noninferiority margin chosen was not sufficiently narrow to exclude excess mortality.60-62 Studies where mortality is an outcome of interest should provide an explicit rationale for the choice of noninferiority margin in specific patient populations. Moreover, “the choice of margin should be independent of considerations of power as the size of the clinically important difference is not altered by the size of the study.”63

Weakness 4. Reporting and Interpreting Nonsignificant Differences

A common statistical error in the medical literature is the conclusion that nonsignificant differences (p>0.05) are “similar.”64 Most studies included in our systematic review were insufficiently powered to address mortality outcomes. Of the five studies32-35,39 of antibiotic discontinuation in ICU patients, only the Bouadma et al. study39 was powered to show noninferiority in mortality between procalcitonin-guided and control group antibiotic therapy. However, three32,33,35 of the four32-35 remaining studies have erroneously concluded that mortality was similar between procalcitonin and control group guided antibiotic therapy based on observed nonsignificant differences in mortality. These three studies reported results as follows: “…without any adverse effects on outcome…,”33 “…a similar mortality were observed in procalcitonin and control groups…,”32 and “…absence of differences in overall mortality suggest that procalcitonin guided antibiotic reduction is not associated with a worse outcome in VAP…”35 Clearly stating in the abstract that the study was not powered to detect a difference in mortality would provide a more accurate reporting of the results.

PubReader format: click here to try


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (2.1M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...