Close

Search

Abstract

Objectives

To summarize available evidence on the effect of continuous infusion (CoI) of vancomycin compared with intermittent infusion (InI) in adult patients with Gram-positive infections.

Methods

MEDLINE, EMBASE and Cochrane databases were searched. Randomized clinical trials (RCTs) and observational studies that comparatively assessed CoI and InI of vancomycin in terms of mortality, clinical cure, toxicity rates and serum drug exposure [trough concentration (C min) for InI and steady-state concentration (C ss) for CoI; area under the curve at 24 h (AUC24) for both] were included. Meta-analysis was conducted combining and analysing the relative risk (RR) and computing a summary RR of the effects with 95% confidence interval (CI). The standardized mean difference was calculated for continuous outcomes. The I 2 test was calculated to assess heterogeneity across studies.

Results

One RCT and five observational studies were included in the analysis. Compared with InI, CoI of vancomycin was associated with a significantly lower risk of nephrotoxicity (RR 0.6, 95% CI 0.4–0.9, P= 0.02; I 2= 0). Overall mortality was not different between the two groups (RR 1.03, 95% CI 0.7–1.6, P= 0.9; I 2= 0).

Conclusions

Our meta-analysis suggests that administration of vancomycin for the treatment of Gram-positive infections by CoI is associated with a significantly lower risk of nephrotoxicity when compared with InI of the drug. RCTs are needed to define the impact on mortality rate and on the pharmacodynamic activity in terms of AUC/MIC ratio.

Introduction

Vancomycin has for a long time been considered the gold standard for the therapy of methicillin-resistant Staphylococcus aureus (MRSA) infections. The just recently released guidelines from the Infectious Diseases Society of America (IDSA) confirm the prominent role of the drug in the treatment of these infections. 1 Notably, the most recently approved antibiotics—linezolid, daptomycin and tigecycline—did not show a significant superiority for clinical cure rate of MRSA infections. 2–8

However, several issues affect vancomycin use in daily clinical practice. First, clinical failure in patients with severe MRSA infections has been increasingly reported in recent years. 9–11 There is still not definitive evidence on the optimal dosage of vancomycin and the impact of pharmacokinetic and pharmacodynamic parameters on patient outcome. In vitro data indicate that the time above the MIC is the most important pharmacodynamic parameter for its efficacy. 12,13 Experimental animal models 14 and human data from patients with MRSA infections showed some concentration-dependent activity as well. 15 The most efficacious method of administration of vancomycin has been debated for a long time. Previous studies have shown that continuous infusion (CoI) of vancomycin may enable faster and more consistent attainment of therapeutic serum concentrations of antibiotic compared with intermittent infusion (InI) and that CoI was a protective factor for intensive care unit (ICU) mortality in patients with MRSA ventilator-associated pneumonia. 16–18 The consensus review from the American Society of Health-System Pharmacists (ASHP), the IDSA and the Society of Infectious Diseases Pharmacists (SIDP) for therapeutic monitoring of vancomycin and the IDSA guidelines for the treatment of MRSA infections do not recommend the use of a CoI regimen, since an improvement in patient outcome versus intermittent dosing is considered unlikely. 1,19 Nevertheless, many clinicians prefer to administer vancomycin by CoI. Published studies comparing the effectiveness and the tolerability of CoI versus InI of vancomycin showed inconclusive results. 20–29

The main aim of this systematic review was to summarize available evidence on the effect of CoI of vancomycin compared with InI in adult patients with infections due to Gram-positive bacteria.

Methods

Methods of the analysis and inclusion criteria were specified in advance and documented in a protocol.

Article identification

Published articles (from January 1956 to May 2011) reporting the use of CoI of vancomycin in human patients were identified through computerized literature searches using MEDLINE (National Library of Medicine, Bethesda, MD, USA), EMBASE and Cochrane databases and by reviewing the references of retrieved articles. Index search terms included the medical subjects heading 'vancomycin' and 'continuous' or 'dosing' or 'intermittent' or 'infusion' or 'discontinuous' or 'administration'. No restriction of languages was applied. No attempt was made to obtain information about unpublished studies. Reviewed articles were maintained in a master log and any reason for exclusion from analysis was documented in the rejected log.

Eligibility criteria

Types of studies

Hypothesizing that data from randomized clinical trials (RCTs) were limited, observational studies were considered eligible where RCTs were found not to be available. Reviews, letters, editorials, case reports and studies focusing only on pharmacokinetic/pharmacodynamic parameters were excluded. Studies comparing the effects of CoI and InI of vancomycin on the following outcomes were included: mortality, clinical cure, toxicity rates and vancomycin serum drug exposure.

Types of participants

Adult patients (>18 years old) with Gram-positive infection treated with vancomycin were included.

Outcomes

Primary outcomes of the meta-analysis included nephrotoxicity rate and overall mortality. The clinically assessable population comprised patients who satisfied the criteria for eligibility set in each of the included studies. Secondary outcomes of the meta-analysis included clinical failure, adverse effect rates and vancomycin serum drug exposure [trough concentration (C min) for InI and steady-state concentration (C ss) for CoI; area under the curve at 24 h (AUC24) for both]. We used data conforming to any outcome definitions reported in each study.

Study selection and data extraction

Eligibility assessment and extraction of data were performed independently by two investigators (M. A. C. and E. G.). Each investigator was blinded to the other investigator's data extraction. In case of disagreement between the two reviewers, a third reviewer was consulted (E. T.). Data from each study were verified for consistency and accuracy, and were then entered into a standardized computerized database. Abstracted information included author, year of study and publication, country in which the study was conducted, study design, number of patients enrolled, population characteristics (ward of hospitalization and type and aetiology of infection), vancomycin MICs for the bacterial isolates responsible for the included infections, characteristics of vancomycin administration (type of infusion, dosage, administration of bolus, dose adjustment and length of therapy), determination of vancomycin serum concentration (C min for InI and C ss for CoI), AUC24 values, adverse effects, clinical failure and overall mortality.

Quality appraisal

Included studies were appraised for methodological quality independently by two authors (M. A. C. and E. G.) without blinding to journal or study authorship. Discrepancies were resolved by discussion or involvement of a third review author if required.

The following risks of bias in randomized trials were assessed, according to the criteria developed by the Cochrane Effective Practice and Organisation of Care (EPOC) group: 30 generation and concealment of allocation; baseline measurement; baseline characteristics; incomplete outcome data; blinded assessment of primary outcomes; protection against contamination and selective outcome reporting.

The quality of observational studies was assessed using the Newcastle-Ottawa scales. 31 In addition, the presence of performance bias (concerning the quality of the information regarding who received what intervention) and of detection bias (concerning correct assessment of outcomes) was assessed in the observational studies.

Statistical analysis

Systematic review was conducted combining and analysing the relative risk (RR) and computing a summary RR of the effects with 95% confidence interval (CI). The standardized mean difference (SDM) was calculated for continuous outcomes. The analysis was performed using the inverse variance fixed effects method, since fixed and random models computed the same overall estimate and 95% CI, showing the absence of 'extra' variability across combined studies. Sensitivity analysis was performed stratifying for type of Gram-positive infection, clinical setting, administration of bolus of vancomycin and dosage adjustment according to vancomycin serum concentration. Meta-analysis was done if more than two studies reporting data on the same outcomes were available. The I 2 test was calculated to assess whether results varied no more than might have been expected by the play of chance (random sampling); significant heterogeneity was considered for I 2 >50%. The publication bias was measured by the Begg funnel plot 32 and the Egger test. 33

Analysis was performed using the software program Intercooled Stata (release 9.0; Stata Statistical Software, College Station, TX, USA).

Results

Figure1 shows the selection process of studies included in the meta-analysis. Six studies were included, 20–25 comprising 443 patients treated with vancomycin: 267 by CoI and 176 by InI.

Figure 1.

Flow chart depicting the selection process of studies included in the meta-analysis.

Flow chart depicting the selection process of studies included in the meta-analysis.

Study description

A summary description of the included studies is reported in Table1. All studies included patients treated between 1983 and 2008. Five studies were performed in Europe 20–24 and one in Asia. 25 Table2 details vancomycin administration in the included studies. The vancomycin MIC values were reported in two studies 22,23 that used the agar dilution method and the Etest, respectively. Vuagnat et al. 23 reported that MICs of bacterial isolates from patients treated with InI were 1.8 ± 0.6 mg/L (range 1–3 mg/L) and from those treated with CoI were 1.9 ± 0.8 mg/L (range 1–4 mg/L). In the RCT, vancomycin MICs were measured in 40 randomly selected strains and were 1 mg/L in 28% and 36% of strains in the InI and CoI group, respectively, and 2 mg/L in 72% and 50% of strains in the InI and CoI group, respectively. 22

Table 1.

Characteristics of studies included in the systematic review

Year Design Setting Type of infection Pathogens No. of patients
 No. of patients concurrently treated with other antibiotics
 Nephrotoxicity definition Mean CREA at baseline (μM)
CoI InI CoI InI CoI InI
1995 20 PrCa ICU bacteraemia/pneumonia MRSA 13 13 NR NR riseb in CREA of 44.2 μM or morec or a rise of 88.4 μM or mored 113 143
1998 21 ReC ICU bacteraemia/pneumonia MRSA/MRCNS 11 14 11 MON/AG 14 MON/AG NA NR NR
2001 22 RCT ICU severe hospital acquired MRSA/MRCNS 61 58 13 FA; 6 AG 13 FA; 16 AG 50% increase in CREAe 98 88
2004 23 PrC medical/surgical ward osteomyelitis MRSA/MRCNS 23 21 5 RIF; 4 CIP 9 RIF; 2 CIP 50% increase in CREAe 84.6 84.7
2009 24 ReC cardio-surgical ICU post-cardiac surgery Gram-positive 119 30 31 CAR/CEPH; 14 AG 8 CAR/CEPH; 3 AG increase in CREAf of at least 0.3 mg/dL, a percentage increase in CREA of at least 50% or a reduction in urine output 79 79
2009 25 ReCg OPAT unit all Gram-positive 40 40 NR NR 50% increase in CREAe 74.8 75.1
Year Design Setting Type of infection Pathogens No. of patients
 No. of patients concurrently treated with other antibiotics
 Nephrotoxicity definition Mean CREA at baseline (μM)
CoI InI CoI InI CoI InI
1995 20 PrCa ICU bacteraemia/pneumonia MRSA 13 13 NR NR riseb in CREA of 44.2 μM or morec or a rise of 88.4 μM or mored 113 143
1998 21 ReC ICU bacteraemia/pneumonia MRSA/MRCNS 11 14 11 MON/AG 14 MON/AG NA NR NR
2001 22 RCT ICU severe hospital acquired MRSA/MRCNS 61 58 13 FA; 6 AG 13 FA; 16 AG 50% increase in CREAe 98 88
2004 23 PrC medical/surgical ward osteomyelitis MRSA/MRCNS 23 21 5 RIF; 4 CIP 9 RIF; 2 CIP 50% increase in CREAe 84.6 84.7
2009 24 ReC cardio-surgical ICU post-cardiac surgery Gram-positive 119 30 31 CAR/CEPH; 14 AG 8 CAR/CEPH; 3 AG increase in CREAf of at least 0.3 mg/dL, a percentage increase in CREA of at least 50% or a reduction in urine output 79 79
2009 25 ReCg OPAT unit all Gram-positive 40 40 NR NR 50% increase in CREAe 74.8 75.1

AG; aminoglycosides; C, cohort; CAR; carbapenems; CEPH, cephalosporins; CIP, ciprofloxacin; CREA, serum creatinine; FA, fusidic acid; MON, monobactams; MRCNS; methicillin-resistant coagulase-negative staphylococci; NA, not applicable; NR, not reported; OPAT, outpatient parenteral antimicrobial therapy; Pr, prospective; Re, retrospective; RIF, rifampicin.

aPatients receiving CoI were matched with historical patients who received InI; matching criteria were site of infection, sex, body weight, severity of illness, duration of therapy, value of serum creatinine concentration before vancomycin therapy and age.

bThe rise was determined by subtracting the initial creatinine concentration from the highest creatinine concentration measured during therapy or within 48 h after therapy.

cIf the initial creatinine was less than 3 mg/100 mL (265.2 μM).

dIf the initial creatinine was 3 mg/100 mL or above.

eFrom the day treatment was started to the end of treatment.

fAn abrupt (within 48 h) reduction in kidney function.

gPatients from a cohort study were matched based on the propensity score estimating the probability of being given CoI of vancomycin. Factors used in the propensity score matching process were diabetes mellitus, baseline serum creatinine and MRSA aetiology.

Table 1.

Characteristics of studies included in the systematic review

Year Design Setting Type of infection Pathogens No. of patients
 No. of patients concurrently treated with other antibiotics
 Nephrotoxicity definition Mean CREA at baseline (μM)
CoI InI CoI InI CoI InI
1995 20 PrCa ICU bacteraemia/pneumonia MRSA 13 13 NR NR riseb in CREA of 44.2 μM or morec or a rise of 88.4 μM or mored 113 143
1998 21 ReC ICU bacteraemia/pneumonia MRSA/MRCNS 11 14 11 MON/AG 14 MON/AG NA NR NR
2001 22 RCT ICU severe hospital acquired MRSA/MRCNS 61 58 13 FA; 6 AG 13 FA; 16 AG 50% increase in CREAe 98 88
2004 23 PrC medical/surgical ward osteomyelitis MRSA/MRCNS 23 21 5 RIF; 4 CIP 9 RIF; 2 CIP 50% increase in CREAe 84.6 84.7
2009 24 ReC cardio-surgical ICU post-cardiac surgery Gram-positive 119 30 31 CAR/CEPH; 14 AG 8 CAR/CEPH; 3 AG increase in CREAf of at least 0.3 mg/dL, a percentage increase in CREA of at least 50% or a reduction in urine output 79 79
2009 25 ReCg OPAT unit all Gram-positive 40 40 NR NR 50% increase in CREAe 74.8 75.1
Year Design Setting Type of infection Pathogens No. of patients
 No. of patients concurrently treated with other antibiotics
 Nephrotoxicity definition Mean CREA at baseline (μM)
CoI InI CoI InI CoI InI
1995 20 PrCa ICU bacteraemia/pneumonia MRSA 13 13 NR NR riseb in CREA of 44.2 μM or morec or a rise of 88.4 μM or mored 113 143
1998 21 ReC ICU bacteraemia/pneumonia MRSA/MRCNS 11 14 11 MON/AG 14 MON/AG NA NR NR
2001 22 RCT ICU severe hospital acquired MRSA/MRCNS 61 58 13 FA; 6 AG 13 FA; 16 AG 50% increase in CREAe 98 88
2004 23 PrC medical/surgical ward osteomyelitis MRSA/MRCNS 23 21 5 RIF; 4 CIP 9 RIF; 2 CIP 50% increase in CREAe 84.6 84.7
2009 24 ReC cardio-surgical ICU post-cardiac surgery Gram-positive 119 30 31 CAR/CEPH; 14 AG 8 CAR/CEPH; 3 AG increase in CREAf of at least 0.3 mg/dL, a percentage increase in CREA of at least 50% or a reduction in urine output 79 79
2009 25 ReCg OPAT unit all Gram-positive 40 40 NR NR 50% increase in CREAe 74.8 75.1

AG; aminoglycosides; C, cohort; CAR; carbapenems; CEPH, cephalosporins; CIP, ciprofloxacin; CREA, serum creatinine; FA, fusidic acid; MON, monobactams; MRCNS; methicillin-resistant coagulase-negative staphylococci; NA, not applicable; NR, not reported; OPAT, outpatient parenteral antimicrobial therapy; Pr, prospective; Re, retrospective; RIF, rifampicin.

aPatients receiving CoI were matched with historical patients who received InI; matching criteria were site of infection, sex, body weight, severity of illness, duration of therapy, value of serum creatinine concentration before vancomycin therapy and age.

bThe rise was determined by subtracting the initial creatinine concentration from the highest creatinine concentration measured during therapy or within 48 h after therapy.

cIf the initial creatinine was less than 3 mg/100 mL (265.2 μM).

dIf the initial creatinine was 3 mg/100 mL or above.

eFrom the day treatment was started to the end of treatment.

fAn abrupt (within 48 h) reduction in kidney function.

gPatients from a cohort study were matched based on the propensity score estimating the probability of being given CoI of vancomycin. Factors used in the propensity score matching process were diabetes mellitus, baseline serum creatinine and MRSA aetiology.

Table 2.

Characteristics of vancomycin administration in the included studies

Reference Loading dose Vancomycin dosage
 Target vancomycin serum concentration
 Mean length (days) of vancomycin treatment
CoI InI CoI InI CoI InI
Wysocki et al. 20 15 mg/kg 30 mg/kg/day 15 mg/kg bid C ss 20–30 mg/L C max 20–40 mg/L and C min 5–10 mg/L 16 16
Di Filippo et al. 21 500 mg 83 mg/h 500 mg qid NA NA 6 6
Wysocki et al. 22 15 mg/kg 30 mg/kg/day 15 mg/kg bid C ss 20–25 mg/L C min 10–15 mg/L 13 14
Vuagnat et al. 23 20 mg/kg 40 mg/kg/day 20 mg/kg bid C ss 20–25 mg/L C max < 50 mg/L and C min 20–25 mg/L 101 66
Hutschala et al. 24 20 mg/kg 15 mg/kg/h according to target C min a C ss 20–25 mg/L C min 15 mg/L 9 9
Ingram et al. 25 NR at discretion of the attending physician at discretion of the attending physician NA NA 22 20
Reference Loading dose Vancomycin dosage
 Target vancomycin serum concentration
 Mean length (days) of vancomycin treatment
CoI InI CoI InI CoI InI
Wysocki et al. 20 15 mg/kg 30 mg/kg/day 15 mg/kg bid C ss 20–30 mg/L C max 20–40 mg/L and C min 5–10 mg/L 16 16
Di Filippo et al. 21 500 mg 83 mg/h 500 mg qid NA NA 6 6
Wysocki et al. 22 15 mg/kg 30 mg/kg/day 15 mg/kg bid C ss 20–25 mg/L C min 10–15 mg/L 13 14
Vuagnat et al. 23 20 mg/kg 40 mg/kg/day 20 mg/kg bid C ss 20–25 mg/L C max < 50 mg/L and C min 20–25 mg/L 101 66
Hutschala et al. 24 20 mg/kg 15 mg/kg/h according to target C min a C ss 20–25 mg/L C min 15 mg/L 9 9
Ingram et al. 25 NR at discretion of the attending physician at discretion of the attending physician NA NA 22 20

bid, twice a day; C max, vancomycin peak concentration; C min, vancomycin trough concentration; C ss, vancomycin steady-state concentration; NA, not applicable; qid, four times a day.

aAuthors did not report a standard dose for InI but stated that dosage was adjusted according to serum creatinine concentration and vancomycin concentration. The daily doses of infused vancomycin were comparable between treatment groups. Eighty-three percent of patients in the InI group received a single daily dose to reach the target vancomycin concentration.

Table 2.

Characteristics of vancomycin administration in the included studies

Reference Loading dose Vancomycin dosage
 Target vancomycin serum concentration
 Mean length (days) of vancomycin treatment
CoI InI CoI InI CoI InI
Wysocki et al. 20 15 mg/kg 30 mg/kg/day 15 mg/kg bid C ss 20–30 mg/L C max 20–40 mg/L and C min 5–10 mg/L 16 16
Di Filippo et al. 21 500 mg 83 mg/h 500 mg qid NA NA 6 6
Wysocki et al. 22 15 mg/kg 30 mg/kg/day 15 mg/kg bid C ss 20–25 mg/L C min 10–15 mg/L 13 14
Vuagnat et al. 23 20 mg/kg 40 mg/kg/day 20 mg/kg bid C ss 20–25 mg/L C max < 50 mg/L and C min 20–25 mg/L 101 66
Hutschala et al. 24 20 mg/kg 15 mg/kg/h according to target C min a C ss 20–25 mg/L C min 15 mg/L 9 9
Ingram et al. 25 NR at discretion of the attending physician at discretion of the attending physician NA NA 22 20
Reference Loading dose Vancomycin dosage
 Target vancomycin serum concentration
 Mean length (days) of vancomycin treatment
CoI InI CoI InI CoI InI
Wysocki et al. 20 15 mg/kg 30 mg/kg/day 15 mg/kg bid C ss 20–30 mg/L C max 20–40 mg/L and C min 5–10 mg/L 16 16
Di Filippo et al. 21 500 mg 83 mg/h 500 mg qid NA NA 6 6
Wysocki et al. 22 15 mg/kg 30 mg/kg/day 15 mg/kg bid C ss 20–25 mg/L C min 10–15 mg/L 13 14
Vuagnat et al. 23 20 mg/kg 40 mg/kg/day 20 mg/kg bid C ss 20–25 mg/L C max < 50 mg/L and C min 20–25 mg/L 101 66
Hutschala et al. 24 20 mg/kg 15 mg/kg/h according to target C min a C ss 20–25 mg/L C min 15 mg/L 9 9
Ingram et al. 25 NR at discretion of the attending physician at discretion of the attending physician NA NA 22 20

bid, twice a day; C max, vancomycin peak concentration; C min, vancomycin trough concentration; C ss, vancomycin steady-state concentration; NA, not applicable; qid, four times a day.

aAuthors did not report a standard dose for InI but stated that dosage was adjusted according to serum creatinine concentration and vancomycin concentration. The daily doses of infused vancomycin were comparable between treatment groups. Eighty-three percent of patients in the InI group received a single daily dose to reach the target vancomycin concentration.

Quality of included studies

Regarding the risk of bias in the RCT, the method used to generate the allocation sequence was adequate, whereas the allocation concealment was not described. Baseline measurement and characteristics were provided and were similar between groups. Missing outcomes were not reported. Primary outcome was blinded assessed. Protection against contamination was not described. The trial was free from selective outcome reporting.

Table3 shows the quality appraisal of the included studies. Regarding other risks of bias in the observational studies, detection bias could have occurred, considering that no blinded assessment of outcome was done in all studies; however, there was no difference among the two groups of treatment.

Table 3.

Quality appraisal of observational studies (indicators from Newcastle-Ottawa scale 31)

References Quality indicators
1a 2b 3c 4d 5Ae 5Bf 6g 7h 8i
Wysocki et al. 20 yes partialj yes no yes yes yes yes NR
Di Filippo et al. 21,k yes yes yes no no no yes yes NR
Vuagnat et al. 23 selected group yes yes yes no no yes yes no (61%)
Hutschala et al. 24 selected group yes yes no no no yes yes NR
Ingram et al. 25 selected group yes yes no yes yes yes yes NR
References Quality indicators
1a 2b 3c 4d 5Ae 5Bf 6g 7h 8i
Wysocki et al. 20 yes partialj yes no yes yes yes yes NR
Di Filippo et al. 21,k yes yes yes no no no yes yes NR
Vuagnat et al. 23 selected group yes yes yes no no yes yes no (61%)
Hutschala et al. 24 selected group yes yes no no no yes yes NR
Ingram et al. 25 selected group yes yes no yes yes yes yes NR

NR, not reported.

aIndicates exposed cohort truly representative.

bNon-exposed cohort drawn from the same community.

cAscertainment of exposure from a secure record.

dOutcome of interest not present at start of study.

eCohorts comparable on basis of site and aetiology of infection.

fCohorts comparable on other factors.

gAssessment of outcome from record linkage or independent blind assessment.

hFollow-up long enough for outcomes to occur.

iComplete accounting for cohorts.

jSame hospital, but not the same period of hospitalization.

kThis study was not included in any pooled analysis.

Table 3.

Quality appraisal of observational studies (indicators from Newcastle-Ottawa scale 31)

References Quality indicators
1a 2b 3c 4d 5Ae 5Bf 6g 7h 8i
Wysocki et al. 20 yes partialj yes no yes yes yes yes NR
Di Filippo et al. 21,k yes yes yes no no no yes yes NR
Vuagnat et al. 23 selected group yes yes yes no no yes yes no (61%)
Hutschala et al. 24 selected group yes yes no no no yes yes NR
Ingram et al. 25 selected group yes yes no yes yes yes yes NR
References Quality indicators
1a 2b 3c 4d 5Ae 5Bf 6g 7h 8i
Wysocki et al. 20 yes partialj yes no yes yes yes yes NR
Di Filippo et al. 21,k yes yes yes no no no yes yes NR
Vuagnat et al. 23 selected group yes yes yes no no yes yes no (61%)
Hutschala et al. 24 selected group yes yes no no no yes yes NR
Ingram et al. 25 selected group yes yes no yes yes yes yes NR

NR, not reported.

aIndicates exposed cohort truly representative.

bNon-exposed cohort drawn from the same community.

cAscertainment of exposure from a secure record.

dOutcome of interest not present at start of study.

eCohorts comparable on basis of site and aetiology of infection.

fCohorts comparable on other factors.

gAssessment of outcome from record linkage or independent blind assessment.

hFollow-up long enough for outcomes to occur.

iComplete accounting for cohorts.

jSame hospital, but not the same period of hospitalization.

kThis study was not included in any pooled analysis.

Nephrotoxicity

Five of six studies were assessable for the nephrotoxicity risk. 20,22–25 Compared with InI, administration by CoI significantly reduced the risk of nephrotoxicity of vancomycin (RR 0.6, 95% CI 0.4–0.9; P= 0.02; Figure2). No significant heterogeneity between the studies was documented (I 2= 0).

Figure 2.

Forest plot summary (fixed effect) of the unadjusted RR of the studies included in the meta-analysis comparing nephrotoxicity rates in patients treated with CoI versus InI of vancomycin.

Forest plot summary (fixed effect) of the unadjusted RR of the studies included in the meta-analysis comparing nephrotoxicity rates in patients treated with CoI versus InI of vancomycin.

Regarding the severity of nephrotoxicity, two studies reported data regarding patients who required dialysis. In the RCT, dialysis was required for 5% (3/58) of patients in the InI group and for 10% (6/61) in the CoI group. 22 Hutschala et al. 24 reported that dialysis was required for 30% (9/30) of patients in the InI group and for 24% (28/119) in the CoI group. The difference between groups was not statistically significant in both studies.

Mortality

Four of six studies were included in the evaluation of the overall mortality. 20,22–24 The combined RR for the overall mortality in patients treated with CoI versus InI was 1.03 (95% CI 0.7–1.6; P= 0.9; Figure3). There was no significant heterogeneity between studies (I 2= 0). After excluding from the analysis one study that did not differentiate patients who died from those lost to follow-up, 23 the combined RR did not differ. CoI of vancomycin did not seem to be effective in significantly reducing the mortality rate, neither among patients with MRSA infections 20,22,23 (total number of included patients, 189; RR 1.2, 95% CI 0.6–2.2; P= 0.6), nor among ICU patients 20,22,24 (total number of included patients, 193; RR 1.03, 95% CI 0.7–1.6; P= 0.9).

Figure 3.

Forest plot summary (fixed effect) of the unadjusted RR of the studies included in the meta-analysis comparing overall mortality rates in patients treated with CoI versus InI of vancomycin.

Forest plot summary (fixed effect) of the unadjusted RR of the studies included in the meta-analysis comparing overall mortality rates in patients treated with CoI versus InI of vancomycin.

Vancomycin exposure

All studies 20–25 analysed the C min and the C ss (Table4). Due to the high statistical heterogeneity between studies (I 2= 90%), the meta-analysis was not carried out. The causes of heterogeneity were not explored due to the low number of included studies.

Table 4.

Vancomycin serum drug exposure values in the included studies

Reference Vancomycin serum concentration (mg/L), ±SD
 AUC24 (mg/L/h), ±SD
CoI (C ss) InI (C min) CoI InI
Wysocki et al. 20 24 ± 6 6 ± 8 — —
Di Filippo et al. 21 24 ± 4 30 ± 6 — —
Wysocki et al. 22 24 ± 8 15 ± 9 577 ± 120 653 ± 232
Vuagnat et al. 23 26 ± 6 22 ± 9 — —
Hutschala et al. 24 25 ± 4 17 ± 5 529 ± 98 612 ± 213
Ingram et al. 25 14 ± 6 10 ± 5 — —
Reference Vancomycin serum concentration (mg/L), ±SD
 AUC24 (mg/L/h), ±SD
CoI (C ss) InI (C min) CoI InI
Wysocki et al. 20 24 ± 6 6 ± 8 — —
Di Filippo et al. 21 24 ± 4 30 ± 6 — —
Wysocki et al. 22 24 ± 8 15 ± 9 577 ± 120 653 ± 232
Vuagnat et al. 23 26 ± 6 22 ± 9 — —
Hutschala et al. 24 25 ± 4 17 ± 5 529 ± 98 612 ± 213
Ingram et al. 25 14 ± 6 10 ± 5 — —

AUC24, area under the serum concentration–time curve over 24 h; C min, vancomycin trough concentration; C ss, vancomycin steady-state concentration.

Table 4.

Vancomycin serum drug exposure values in the included studies

Reference Vancomycin serum concentration (mg/L), ±SD
 AUC24 (mg/L/h), ±SD
CoI (C ss) InI (C min) CoI InI
Wysocki et al. 20 24 ± 6 6 ± 8 — —
Di Filippo et al. 21 24 ± 4 30 ± 6 — —
Wysocki et al. 22 24 ± 8 15 ± 9 577 ± 120 653 ± 232
Vuagnat et al. 23 26 ± 6 22 ± 9 — —
Hutschala et al. 24 25 ± 4 17 ± 5 529 ± 98 612 ± 213
Ingram et al. 25 14 ± 6 10 ± 5 — —
Reference Vancomycin serum concentration (mg/L), ±SD
 AUC24 (mg/L/h), ±SD
CoI (C ss) InI (C min) CoI InI
Wysocki et al. 20 24 ± 6 6 ± 8 — —
Di Filippo et al. 21 24 ± 4 30 ± 6 — —
Wysocki et al. 22 24 ± 8 15 ± 9 577 ± 120 653 ± 232
Vuagnat et al. 23 26 ± 6 22 ± 9 — —
Hutschala et al. 24 25 ± 4 17 ± 5 529 ± 98 612 ± 213
Ingram et al. 25 14 ± 6 10 ± 5 — —

AUC24, area under the serum concentration–time curve over 24 h; C min, vancomycin trough concentration; C ss, vancomycin steady-state concentration.

The AUC24 values were reported in two of six studies. 22,24 Mean values were comparable in the two treatment groups (Table4).

Treatment failure

Only two studies provided treatment failure rates in the two groups of infusion. 22,23 Both reported no significant difference between the two groups.

Adverse effects

Pooled analysis was not performed, given the lack of complete data on adverse effects and the heterogeneity of definitions. In the Vuagnat et al. 23 study, adverse drug reactions leading to withdrawal of vancomycin therapy were significantly more frequent in the InI group. Reported adverse effects were acute renal injury, allergic reaction, severe neutropenia, catheter phlebitis and severe depression in the group receiving InI of vancomycin, whereas only catheter phlebitis was reported in the group receiving CoI. 23 In two studies it was reported that red man syndrome occurred only in few patients under InI, even if the difference was not statistically significant. 22,24

Analysis of publication bias

The Begg's funnel plot and the Egger test indicated that there was no evidence of publication bias.

Discussion

Our meta-analysis shows that CoI of vancomycin, when compared with InI under the same daily dosage, was associated with a significantly lower risk of drug-related nephrotoxicity in patients treated for Gram-positive infections.

Renal toxicity, although relatively rare, is one of most dreaded adverse effects of vancomycin and has been associated with an increased risk of bad outcomes. 34 The lower risk of nephrotoxicity with CoI is probably due to the fact that the desired steady-state concentrations may be attained with lower daily doses than those needed to achieve a similar trough with InI, 17,22 thus avoiding the potential risk deriving from excessive drug exposure in terms of AUC. 35 Indeed, lower variability in plasma concentrations when vancomycin is administered continuously has been described. 23,28

Our meta-analysis was unable to demonstrate a significant difference in mortality rates in patients with Gram-positive infections treated with CoI versus InI of vancomycin. Major pitfalls for the analysis of mortality were the assessment of overall mortality instead of infection-related mortality (not analysed in all studies) and the heterogeneity of aetiology, site and severity of infections.

Notably, heterogeneity of definitions and lack of data in studies did not allow us to carry out a meta-analysis on the impact of the method of vancomycin administration on the serum vancomycin concentration, treatment failure and adverse effects rates.

The consensus review from ASHP, IDSA and SIDP underlines that vancomycin plasma concentration must be in excess of the bacterial MIC and recommends a plasma vancomycin C min of 15–20 mg/L in S. aureus–complicated infections. 19 Some authors argue that the standard dosing regimen of 15 mg/kg every 12 h does not assure attainment of this goal. 26 In a study in which vancomycin plasma concentration was assessed after 36–48 h of InI and CoI, adequate vancomycin plasma concentrations for the treatment of MRSA infections were observed more frequently in patients treated by CoI. 26 However, as far as the time above the MIC is concerned, a cross-over trial comparing the two ways of infusion in patients with Gram-positive infections found that both regimens resulted in the MIC being exceeded 100% of the time. 28

Analysis of costs was not included in the outcomes of our study, although a difference in hospital costs between the two ways of infusion does exist. Wysocki et al. 22 reported that the 10-day treatment cost per patient, including the cost of serum vancomycin determination, was significantly lower in the CoI group.

Our meta-analysis has several limitations. First, it was possible to include a low number of studies and only one RCT. However, meta-analyses were done including only studies reporting the same outcome, and the quality of observational studies included was fairly good. No significant difference was detected between the two groups of treatment for epidemiological and clinical factors. Moreover, the 'hierarchy' of study design has been debated. A study comparing the findings of separate meta-analyses of RCTs and observational studies assessing the same intervention found that the average results were similar. 36 Second, studies did not differentiate between the minor and major kidney injuries that occurred during vancomycin treatment, and many patients received other antibiotics that could have influenced both clinical outcomes and nephrotoxicity. Further sensitivity analysis could not be performed due to the heterogeneity of treatments.

In conclusion, our meta-analysis showed that administration of vancomycin for the treatment of Gram-positive infections by CoI was associated with a significantly lower risk of nephrotoxicity when compared with the InI of the drug. However, since this evidence is supported by four observational studies and one RCT, it cannot be considered conclusive. The research agenda needs to move to multicentre studies applying randomized allocation of the method of infusion, adequate sample size and standardized methods of vancomycin concentration measures and assessing the impact of the method of vancomycin administration on the mortality rate and on pharmacodynamic activity in terms of the AUC/MIC ratio. Cost-effectiveness analysis is also urgently needed.

Funding

This work was not supported by outside funding. All work was funded by our department research budget.

Transparency declarations

E. T. has served on the advisory board to Sanofi-Aventis and Astellas, and has been on the speakers' bureau for Pfizer and Novartis. F. P. has been a consultant for Astellas, has been a consultant for and on the speakers' bureau for Pfizer and has also been on the speakers' bureau for Gilead, GlaxoSmithKline, Merck Sharp & Dohme, Novartis and Sanofi-Aventis. N. P. has served on the advisory board of Care Fusion and Janssen-Cilag, and has been on the speakers' bureau for MSD, Pfizer, Wyeth, Janssen-Cilag, Johnson & Johnson, Gilead, Novartis, Astellas, Sanofi Aventis and GSK. M. A. C. and E. G. have no conflicts of interest to declare.

References

1

,  ,  , et al.

Clinical practice guidelines by the Infectious Diseases Society of America for the treatment of methicillin-resistant Staphylococcus aureus infections in adults and children

,

Clin Infect Dis

,

2011

, vol.

52

 (pg.

e18

-

55

)

2

,  ,  , et al.

Linezolid versus vancomycin for the treatment of Gram-positive bacterial infections: meta-analysis of randomised controlled trials

,

Int J Antimicrob Agents

,

2010

, vol.

35

 (pg.

3

-

12

)

3

,  ,  .

Linezolid versus glycopeptide antibiotics for the treatment of suspected methicillin-resistant Staphylococcus aureus nosocomial pneumonia: a meta-analysis of randomized controlled trials

,

Chest

,

2011

, vol.

139

 (pg.

1148

-

55

)

4

,  .

Efficacy and safety of linezolid in methicillin-resistant Staphylococcus aureus (MRSA) complicated skin and soft tissue infection (cSSTI): a meta-analysis

,

Curr Med Res Opin

,

2010

, vol.

26

 (pg.

407

-

21

)

5

,  ,  .

Linezolid versus vancomycin for Staphylococcus aureus bacteraemia: pooled analysis of randomized studies

,

J Antimicrob Chemother

,

2005

, vol.

56

 (pg.

923

-

9

)

6

,  ,  , et al.

Effectiveness and safety of tigecycline for the treatment of infectious disease: a systematic review and meta-analysis

,

Antimicrob Agents Chemother

,

2011

, vol.

55

 (pg.

1162

-

72

)

7

,  ,  , et al.

Daptomycin versus other antimicrobial agents for the treatment of skin and soft tissue infections: a meta-analysis

,

Ann Pharmacother

,

2010

, vol.

44

 (pg.

97

-

106

)

8

,  ,  , et al.

Meta-analysis of trials evaluating parenteral antimicrobial therapy for skin and soft tissue infections

,

Clin Infect Dis

,

2010

, vol.

50

 (pg.

1120

-

6

)

9

,  ,  , et al.

Treatment outcomes for serious infections caused by methicillin-resistant Staphylococcus aureus with reduced vancomycin susceptibility

,

Clin Infect Dis

,

2004

, vol.

38

 (pg.

521

-

8

)

10

,  ,  , et al.

Relationship between vancomycin MIC and failure among patients with methicillin-resistant Staphylococcus aureus bacteremia treated with vancomycin

,

Antimicrob Agents Chemother

,

2008

, vol.

52

 (pg.

3315

-

20

)

11

,  ,  , et al.

Relationship of MIC and bactericidal activity to efficacy of vancomycin for treatment of methicillin-resistant Staphylococcus aureus bacteremia

,

J Clin Microbiol

,

2004

, vol.

42

 (pg.

2398

-

402

)

12

,  ,  .

In vitro studies of pharmacodynamic properties of vancomycin against Staphylococcus aureus and Staphylococcus epidermidis

,

Antimicrob Agents Chemother

,

1998

, vol.

42

 (pg.

2739

-

44

)

13

,  ,  .

Analysis of vancomycin population susceptibility profiles, killing activity, and postantibiotic effect against vancomycin-intermediate Staphylococcus aureus

,

Antimicrob Agents Chemother

,

1999

, vol.

43

 (pg.

1914

-

8

)

14

,  ,  , et al.

Pharmacodynamics of glycopeptides in the mouse peritonitis model of Streptococcus pneumoniae or Staphylococcus aureus infection

,

Antimicrob Agents Chemother

,

2000

, vol.

44

 (pg.

1247

-

54

)

15

,  ,  , et al.

Pharmacodynamics of vancomycin and other antimicrobials in patients with Staphylococcus aureus lower respiratory tract infections

,

Clin Pharmacokinet

,

2004

, vol.

43

 (pg.

925

-

42

)

16

,  ,  , et al.

Better outcomes through continuous infusion of time-dependent antibiotics to critically ill patients?

,

Curr Opin Crit Care

,

2008

, vol.

14

 (pg.

390

-

6

)

17

,  .

Should the currently recommended twice-daily dosing still be considered the most appropriate regimen for treating MRSA ventilator-associated pneumonia with vancomycin?

,

Clin Pharmacokinet

,

2008

, vol.

47

 (pg.

147

-

52

)

18

,  ,  , et al.

Pneumonia caused by oxacillin-resistant Staphylococcus aureus treated with glycopeptides

,

Crit Care Med

,

2005

, vol.

33

 (pg.

1983

-

7

)

19

,  ,  , et al.

Therapeutic monitoring of vancomycin in adult patients: a consensus review of the American Society of Health-System Pharmacists, the Infectious Diseases Society of America, and the Society of Infectious Diseases Pharmacists

,

Am J Health Syst Pharm

,

2009

, vol.

66

 (pg.

82

-

98

)

20

,  ,  , et al.

Comparison of continuous with discontinuous intravenous infusion of vancomycin in severe MRSA infections

,

J Antimicrob Chemother

,

1995

, vol.

35

 (pg.

352

-

4

)

21

,  ,  , et al.

Continuous infusion of vancomycin in methicillin-resistant Staphylococcus infection

,

Chemotherapy

,

1998

, vol.

44

 (pg.

63

-

8

)

22

,  ,  , et al.

Continuous versus intermittent infusion of vancomycin in severe staphylococcal infections: prospective multicenter randomized study

,

Antimicrob Agents Chemother

,

2001

, vol.

45

 (pg.

2460

-

7

)

23

,  ,  , et al.

High dose vancomycin for osteomyelitis: continuous vs. intermittent infusion

,

J Clin Pharm Ther

,

2004

, vol.

29

 (pg.

351

-

7

)

24

,  ,  , et al.

Influence of vancomycin on renal function in critically ill patients after cardiac surgery: continuous versus intermittent infusion

,

Anesthesiology

,

2009

, vol.

111

 (pg.

356

-

65

)

25

,  ,  , et al.

Nephrotoxicity of continuous versus intermittent infusion of vancomycin in outpatient parenteral antimicrobial therapy

,

Int J Antimicrob Agents

,

2009

, vol.

34

 (pg.

570

-

4

)

26

,  .

Monitoring of vancomycin serum levels for the treatment of staphylococcal infections

,

Clin Microbiol Infect

,

2006

, vol.

12

 (pg.

92

-

5

)

27

,  ,  , et al.

Comparison of continuous infusion versus intermittent infusion of vancomycin in patients with methicillin-resistant Staphylococcus aureus

,

J Med Assoc Thai

,

2010

, vol.

93

 (pg.

172

-

6

)

28

,  ,  , et al.

Comparison of conventional dosing versus continuous-infusion vancomycin therapy for patients with suspected or documented Gram-positive infections

,

Antimicrob Agents Chemother

,

1996

, vol.

40

 (pg.

696

-

700

)

29

,  ,  , et al.

Comparison of bactericidal activities of intermittent and continuous infusion dosing of vancomycin against methicillin-resistant Staphylococcus aureus and Enterococcus faecalis

,

Pharmacotherapy

,

1998

, vol.

18

 (pg.

1069

-

74

)

30

EPOC Data Collection Checklist

 

31

,  ,  , et al.,

The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses

,

2005

 

32

,  .

Operating characteristics of a rank correlation test for publication bias

,

Biometrics

,

1994

, vol.

50

 (pg.

1088

-

101

)

33

,  ,  , et al.

Bias in meta-analysis detected by a simple, graphical test

,

BMJ

,

1997

, vol.

315

 (pg.

629

-

34

)

34

,  ,  , et al.

Minimal changes of serum creatinine predict prognosis in patients after cardiothoracic surgery: a prospective cohort study

,

J Am Soc Nephrol

,

2004

, vol.

15

 (pg.

1597

-

605

)

35

,  ,  , et al.

Relationship between initial vancomycin concentration-time profile and nephrotoxicity among hospitalized patients

,

Clin Infect Dis

,

2009

, vol.

49

 (pg.

507

-

14

)

36

,  ,  .

Randomized, controlled trials, observational studies, and the hierarchy of research designs

,

N Engl J Med

,

2000

, vol.

342

 (pg.

1887

-

92

)

© The Author 2011. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com

© The Author 2011. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com