3. broad-spectrum empiric therapy that included vanco-
mycin; had Ն1 vancomycin trough level collected
within 96 hours of vancomycin therapy; and had a
baseline serum creatinine level Ͻ2 mg/dL. Patients
were excluded if they had a baseline serum creatinine
level Ն2 mg/dL, no serum creatinine recorded at study
entry, or known history of end-stage renal disease or
dialysis at baseline. We collected data on patient demo-
graphic and baseline characteristics; severity of illness,
including the Acute Physiology and Chronic Health Eval-
uation (APACHE) II score, which provides a general
measure of severity of disease by assigning points based
on initial values of 12 routine physiologic measurements,
age, and previous health status16
; diagnostic procedures;
and treatment, including all antibiotics received from the
time of diagnosis of pneumonia. Patients were followed
up until hospital discharge, death, or 28 days after pneu-
monia diagnosis, whichever occurred first. Laboratory
values were collected during the index hospitalization,
including serum creatinine and vancomycin trough levels.
It was recommended that blood for trough concentra-
tions be drawn ϳ1 hour before planned dosing. As part
of the standard of care at each institution, clinical phar-
macists were actively involved in the care of ICU patients
and assisted with appropriate dosing of antibiotics.
Study Definitions
Baseline serum creatinine was defined as the mean
of the values recorded on days –1 and 0, with day 0
being day of pneumonia diagnosis and, in general, ini-
tiation of antibiotic therapy. Nephrotoxicity was de-
fined as an increase in serum creatinine Ն0.5 mg/dL or
50% above baseline, whichever was greater, in at least 2
consecutive measurements during the period from initia-
tion of vancomycin therapy to 72 hours after completion
of therapy.2,14
Because vancomycin is mainly eliminated
by glomerular filtration, any change in renal function can
affect vancomycin serum concentrations. To account for
this potential confounder, we limited our analysis of van-
comycin trough levels to those recorded within the first
96 hours of initiation of therapy.17
All vancomycin
trough levels documented in the medical record were col-
lected, and the highest concentration in the first 96 hours
was used in the analysis.
Statistical Analysis
Statistics were calculated using SAS 9.2 (SAS Insti-
tute Inc, Cary, North Carolina).18
For bivariate analy-
ses, categorical variables were compared using 2
test
or the Fisher exact test, and continuous variables were
compared using the Student t test and the nonparamet-
ric Wilcoxon rank sum test when appropriate.19
To identify factors independently associated with
nephrotoxicity, multivariate logistic regression analy-
ses were conducted.19
Variables associated with neph-
rotoxicity in the bivariate analysis were considered for
inclusion in the models, requiring P Յ 0.20 for each
term to be included. Kaplan-Meier analyses were used
to examine time to nephrotoxicity according to vanco-
mycin trough levels, with patients stratified into 4
groups according to initial vancomycin trough level as
follows: Ͻ10, 10 to Ͻ15, 15 to 20, and Ͼ20 mg/L.
RESULTS
IMPACT-HAP enrolled 449 ICU patients with HAP,
HCAP, or VAP. We excluded 209 patients from anal-
ysis because they had no baseline serum creatinine re-
corded (n ϭ 38), had a baseline serum creatinine Ն2
mg/dL (n ϭ 79), had a known history of end-stage
renal disease or dialysis at baseline (n ϭ 6), or did not
receive treatment with vancomycin (n ϭ 86). These
criteria were applied hierarchically; although it is pos-
sible some patients had multiple exclusion criteria,
they were only counted once (Figure 1).
Of 240 evaluable patients, 188 were treated with
vancomycin and had sufficient clinical and laboratory
follow-up information for analysis. In these 188 pa-
tients, 63% were male; the mean (SD) age was 58.5
(17.2) years; and the mean APACHE II score was 19.4
(6.4). Nephrotoxicity during the period from initiation of
vancomycin therapy to 72 hours after completion of ther-
apy occurred in 29 of 188 vancomycin-treated patients
(15.4%). The demographic and clinical variables signifi-
cantly associated with the occurrence of nephrotoxicity
in bivariate analysis were higher median vancomycin
trough level, initial vancomycin trough level Ն15 mg/L,
concomitant aminoglycoside administration, and shorter
ICU stay before diagnosis of pneumonia (Table I).
In the multivariate logistic regression analysis, vari-
ables independently associated with nephrotoxicity in
vancomycin-treated patients were initial vancomycin
trough level Ն15 mg/L (odds ratio [OR], 5.2 [95% CI,
1.9–13.9]; P ϭ 0.001), concomitant administration of
aminoglycosides (OR, 2.67 [95% CI, 1.09–6.54]; P ϭ
0.03), and duration of vancomycin therapy as a con-
tinuous variable (OR for each additional treatment
day, 1.12 [95% CI, 1.02–1.23]; P ϭ 0.02; Table II).
E.L. Cano et al.
January 2012 151
4. A relationship was observed between vancomycin
trough levels and the occurrence of nephrotoxicity.
Nephrotoxicity significantly increased as a function of
the initial vancomycin serum trough concentration,
rising from 7.0% at a trough Ͻ10 mg/L to 34.0% at
Ͼ20 mg/L (P ϭ 0.001; Figure 2). Kaplan-Meier analysis
revealed a significant difference in mean time to nephro-
toxicity when patients were stratified according to initial
vancomycin trough level (P ϭ 0.0003; Figure 3). Mean
time to nephrotoxicity decreased from 8.8 days at vanco-
mycin trough levels Ͻ15 mg/L to 7.4 days at Ͼ20 mg/L.
Patients who developed nephrotoxicity had a longer
ICU stay (median [interquartile range (IQR)], 17 [9–
26] vs 12 [6–20] days; P ϭ 0.03) and seemed to have a
longer hospital stay (median [IQR], 20 [14–30] vs 15
[9–26] days; P ϭ 0.06) after initiation of antibiotics for
pneumonia than those patients without nephrotoxic-
ity, although the difference in hospital stay did not
reach statistical significance. Between-group differ-
ences in mortality at 28 days were not significant
(33.3% vs 25.3%; P ϭ 0.48).
In an effort to address reverse causality (ie, the
potential confounding effect of acute renal injury
already present at pneumonia diagnosis), the bivari-
ate and multivariate analyses were repeated after
eliminating those patients with evidence of acute re-
nal dysfunction on study day 0. This excluded 3
patients presenting with an increase in serum creat-
inine Ն0.5 mg/dL above their previously known
baseline, 10 patients with a Ͼ50% creatinine in-
crease above baseline, and 21 patients for both rea-
sons. In this subgroup, nephrotoxicity was observed
in 18 of 154 patients (11.7%). Demographic and
clinical characteristics were similar to those of the
188 vancomycin-treated patients. In these 154 pa-
tients, 63% were male and the mean age was 58.4
(17.2) years. Because APACHE II includes points for
renal dysfunction, the mean score—not unexpectedly—
Figure 1. Flow of patients through analysis. ESRD ϭ end-stage renal disease; IMPACT-HAP ϭ Improving Medicine
Through Pathway Assessment of Critical Therapy in Hospital-Acquired Pneumonia.
Clinical Therapeutics
152 Volume 34 Number 1
5. was slightly lower in the patients without acute
nephrotoxicity at baseline (18.3 [6.0] vs 19.4 [6.4];
P Ͻ 0.01). In multivariate logistic regression analy-
sis, the only variable independently associated with
nephrotoxicity was duration of vancomycin therapy
(OR, 1.15 [95% CI, 1.03–1.28]; P ϭ 0.01). The OR
for vancomycin trough Ն15 mg/L was 2.93 (95%
CI, 0.96–8.92; P ϭ 0.06).
Table I. Comparison of demographics, comorbid conditions, and clinical characteristics of vancomycin-treated
patients (n ϭ 188) who experienced nephrotoxicity and those who did not.*
Characteristic†
Nephrotoxicity
(n ϭ 29)
No Nephrotoxicity
(n ϭ 159) P
Age, mean (SD), y 59.9 (17.5) 58.3 (17.2) 0.66
Male sex, no. (%) 23 (79.3) 96 (60.4) 0.06
Weight, mean (SD), lb 177.4 (42.0) 187.6 (71.8) 0.29
Vancomycin trough level, median (IQR) 22.5 (17–30) 14 (10.5–19) 0.05
Initial vancomycin trough level, no. (%) 0.001
Ͻ15 mg/L 7 (24.1) 92 (57.9)
Ն15 mg/L 22 (75.9) 67 (42.1)
Length of vancomycin therapy, median (IQR), d 8 (5–12) 7 (4–11) 0.16
Concomitant aminoglycoside use, no. (%) 16 (55.2) 46 (28.9) 0.009
Vascular disease, no. (%) 5 (17.2) 39 (24.5) 0.48
Diabetes mellitus, no. (%) 9 (31.0) 43 (27.0) 0.66
Severity of illness
CPIS, mean (SD) 6.4 (1.9) 6.5 (1.6) 0.89
Severe sepsis, no. (%) 23 (79.3) 127 (79.9) 1.0
APACHE II, mean (SD) 19.9 (7.2) 19.3 (6.2) 0.64
Length of stay before NP diagnosis, median (IQR)
Hospital 5 (1–7) 8 (2–13) 0.09
ICU 2 (0–5) 5 (0–11) 0.05
IQR ϭ interquartile range; CPIS ϭ Clinical Pulmonary Infection Score; APACHE II ϭ Acute Physiology and Chronic Health
Evaluation II; NP ϭ nosocomial pneumonia; ICU ϭ intensive care unit.
*All percentages were based on number of evaluable patients (ie, excludes patients with missing data).
†
The following characteristics were included in the bivariate model but are not shown above because P Ͼ 0.2: race, type of
pneumonia, renal disease, cardiac disease, cardiosystem dysfunction, and mechanical ventilation.
Table II. Multivariate logistic regression model for the occurrence of nephrotoxicity in vancomycin-treated
patients.
Parameter Odds Ratio (95% CI) P
Initial vancomycin trough level Ն15 mg/L 5.2 (1.9–13.9) 0.001
Concomitant aminoglycoside use 2.67 (1.09–6.54) 0.03
Length of vancomycin therapy* 1.12 (1.02–1.23) 0.02
*Odds ratio for each additional day.
E.L. Cano et al.
January 2012 153
6. Because aminoglycoside therapy was found to be
associated with nephrotoxicity by univariate analysis
and remained an independent risk factor for nephro-
toxicity in patients receiving vancomycin, these pa-
tients were further analyzed. Of 188 evaluable patients
in the vancomycin-treated group, 126 (67.0%) did not
receive any aminoglycoside therapy. The mean dura-
tion of aminoglycoside therapy was 1.61 (3.25) days
(range, 0–17 days). Of the 29 patients who received
vancomycin and developed nephrotoxicity, 16 had re-
ceived at least 1 dose of an aminoglycoside.
DISCUSSION
Nephrotoxicity is a frequent complication in ICU pa-
tients. Even modest decreases in renal function are as-
sociated with negative outcomes in critically ill pa-
tients, including increased mortality, hospital length of
stay, duration of mechanical ventilation, and hospital
costs.20–22
The recent recommendations in guideline
documents1,3
and an expert review2
to maintain higher
vancomycin trough levels when this antibiotic is used
as part of broad-spectrum empiric therapy for pneu-
monia stimulated our interest in examining the poten-
tial outcomes and consequences of implementing this
practice in ICU patients with pneumonia.
We found that nephrotoxicity occurred in 15.4% of
188 ICU patients who received vancomycin for the
treatment of HAP, VAP, or HCAP. The frequency of
nephrotoxicity increased as a function of initial vanco-
mycin trough value, and the mean time to nephrotox-
icity decreased accordingly. After adjusting for other
potential confounders, a vancomycin trough concen-
tration Ն15 mg/L was associated with a 5-fold increase
in the risk of nephrotoxicity. Other factors indepen-
dently associated with the occurrence of nephrotoxic-
ity in vancomycin-treated patients included duration
of vancomycin therapy and concomitant administra-
tion of aminoglycoside antibiotics. To rule out the pos-
sibility of reverse causality, we reanalyzed our data
after excluding patients with evidence of acute renal
dysfunction at the time of pneumonia diagnosis. In this
analysis, the duration of vancomycin therapy was as-
sociated with nephrotoxicity. The risk associated with
higher vancomycin trough levels decreased, and the
association with nephrotoxicity failed to meet statis-
tical significance. Therefore, acute renal disease did
not seem to bias our findings in the larger group
analysis.
Our findings are consistent with previous studies
and add new insights about the specific risk for neph-
rotoxicity in ICU patients receiving vancomycin for
HAP, VAP, and HCAP. Previous reports were predom-
inantly single-center studies that examined the inci-
dence of nephrotoxicity in patients with various types
of infections,8–13
with only one study focusing on pa-
tients with pneumonia, specifically HCAP.9
The re-
ported incidence of renal dysfunction associated with
vancomycin use was relatively constant in most re-
ports, ranging from 12% to 19%,8,10,17,23,24
and con-
Figure 3. Kaplan-Meier analysis of time to nephro-
toxicity stratified according to initial van-
comycin trough values (P ϭ 0.0003).
Figure 2. Relationship between the initial vanco-
mycin trough value and the frequency of
nephrotoxicity for 188 vancomycin-
treated patients (P ϭ 0.001).
Clinical Therapeutics
154 Volume 34 Number 1
7. sistent with that in our study. There are 2 outlying
articles that merit further consideration. Matsko et al25
reported the lowest incidence (7% in 299 patients in a
240-bed government teaching hospital). Although
their meeting abstract provided no additional details
about the patient population, they also found a signif-
icant relationship between initial vancomycin troughs
and nephrotoxicity. Jeffres et al9
reported the highest
incidence (43% in 102 patients with HCAP due to
MRSA in a 1200-bed urban teaching hospital), which
was attributed to aggressive vancomycin dosing and
prolonged vancomycin administration.
Importantly, others have reported that vancomycin
was an independent risk factor for nephrotoxicity, es-
pecially when analyzed according to increased serum
concentrations as trough Ն15 mg/L8–13
or Ն22.5 mg/
L,25
steady-state concentration Ͼ28 mg/L,24
highest
trough,23
mean trough,17,23
and duration of ther-
apy.12,13,23,25
Additional independent risk factors, all
potentially present in our IMPACT-HAP population,
included ICU residence11,17
and concomitant use of
aminoglycosides24
or other nephrotoxic agents.11,13,23
Collectively, these studies may have important im-
plications for practical implementation of the current
ATS/IDSA guideline.1
Our results and those of others
question whether vancomycin should be dosed to
achieve target trough levels of 15 to 20 mg/L when
used in ICU patients with HAP, VAP, or HCAP. Since
the ATS/IDSA guideline was published, Jeffres et al26
reported that higher mean vancomycin trough values
and higher mean AUC values did not have favorable
effects on survival in patients with HCAP due to
MRSA. In a subsequent study,9
the same investigators
concluded that aggressive dosing strategies for vanco-
mycin may not offer any advantages in clinical efficacy
and that alternative agents should be considered. Based
on our data and those of others, the benefit of increased
vancomycin dosing to achieve higher troughs should
be weighed against the risk of nephrotoxicity.
Our study has several strengths, including a well-
characterized study population with clinical and lab-
oratory information available for analysis.15,27
For
consistency with other studies, we used a nephrotox-
icity definition that is a reasonable composite from
the literature and matches the definition for vanco-
mycin-induced nephrotoxicity used in the 2009 con-
sensus statement by the American Society of Health-
System Pharmacists, the IDSA, and the Society of
Infectious Diseases Pharmacists.2
This definition is
based on changes in serum creatinine and is similar
to risk level in the RIFLE (Risk, Injury, and Failure;
and Loss, and End Stage Kidney Disease) criteria28
and
the Stage I in the Acute Kidney Injury Network crite-
ria.29
By limiting our analysis to patients’ initial van-
comycin trough values, we improved our capacity to
determine whether a causal exposure–response rela-
tionship exists. To the best of our knowledge, this is the
largest study to date describing incidence of nephro-
toxicity and the role of vancomycin in development of
nephrotoxicity in ICU patients with HAP, VAP, and
HCAP.
Our study also has important limitations. IMPACT-
HAP was an observational performance improvement
project, not a randomized controlled trial. As such,
IMPACT-HAP did not dictate prescribing practice;
however, vancomycin trough levels were recom-
mended by the study team in accordance with the
guidelines and managed locally as part of routine care
by clinical pharmacists in participating ICUs. Al-
though this was a retrospective analysis, investigators
completed data-collection forms and validated the in-
formation locally before transferring it via the Internet
to the IMPACT-HAP coordinating center, where it un-
derwent a second validation. Even with the relatively
large study group, some subsets were too small for
detailed analysis, such as a more thorough evaluation
of patients receiving concurrent vancomycin and ami-
noglycoside therapy. The use of inclusion and exclu-
sion criteria may limit the generalizability of our re-
sults. Our patients were adults in multiple ICUs who
had HAP, VAP, or HCAP, an inherently complex
group for analysis; however, they were a more homog-
enous group than those in previous studies that in-
cluded patients with mixed infections in widely varied
settings.8,17,23,24
Although we collected detailed infor-
mation on antibiotic use, we did not have information
on exposure to all nephrotoxic agents, such as intrave-
nous contrast dye. Similarly, although we excluded pa-
tients with evidence of acute renal dysfunction at diag-
nosis, we did not include detailed data on shock and
other ICU-related risk factors for nephrotoxicity.
CONCLUSIONS
Our study provides evidence that nephrotoxicity may
be common among ICU patients with HAP, VAP, or
HCAP who receive broad-spectrum antibiotic therapy
that includes vancomycin. An exposure–response rela-
tionship exists between initial vancomycin trough level
E.L. Cano et al.
January 2012 155
8. and occurrence and mean time to nephrotoxicity. A
vancomycin trough level Ն15 mg/L may be an inde-
pendent risk factor for nephrotoxicity in ICU pneumo-
nia patients. Our findings suggest the need for additional
studies to assess the current guidelines’ recommendation
for vancomycin dosing.
ACKNOWLEDGMENTS
Funding for this study was provided by Pfizer Inc, US
Medical. The University of Louisville Foundation was
responsible for project oversight and distribution of
funds to participating institutions.
The IMPACT-HAP Investigators include the fol-
lowing: Andrea S. Castelblanco and G. Fernando Cu-
billos (University of Miami, Jackson Memorial Hospi-
tal, and Veterans Affairs Medical Center, Miami,
Florida); Carol Moore, Paola Osaki-Kiyan, and Mary
Perri (Henry Ford Health System, Detroit, Michigan);
and Julie E. Mangino, Carol Myers, David Taylor,
Lindsay Pell, and Kari Mount (The Ohio State Univer-
sity, Columbus, Ohio).
Editorial and medical writing support was provided
by Cindy Hamilton from Hamilton House and was
funded by Pfizer Inc.
Drs. Welch, Scerpella, and Ford, investigators who
were employed by the sponsor, participated in the pro-
cesses of study design, data analysis and interpretation,
and contributed to writing and editing of the manu-
script. Drs. Cano, Haque, Cely, Peyrani, Zervos,
Ramirez, and Kett participated in the process of study
design, data analysis and interpretation, and contrib-
uted to writing and editing of the manuscript. Those
with lower levels of responsibilities are listed as
IMPACT-HAP investigators.
CONFLICTS OF INTEREST
Dr. Zervos is a consultant for Astellas and Novartis
and has received honoraria or speaking fees from As-
tellas, Cubist, and Pfizer, and grants from Pfizer, Astel-
las, and Cubist. Dr. Ramirez is a consultant for Pfizer,
Merck, and Cubist and has received honoraria or
speaking fees from Pfizer, Merck, Cubist, and Ortho,
and grants from Pfizer and Cubist. Dr. Kett is a con-
sultant for Pfizer and Astellas and has received hono-
raria or speaking fees from Pfizer, Astellas, and Glaxo
Smith-Kline, and grants from Pfizer and Akers Biosci-
ence. Drs. Welch, Scerpella, and Ford are employees of
Pfizer and own stock and stock options in Pfizer. The
other authors have indicated that they have no other
conflicts of interest with regard to the content of the
article.
REFERENCES
1. American Thoracic Society and the Infectious Diseases
Society of America. Guidelines for the management of
adults with hospital-acquired, ventilator-associated, and
healthcare-associated pneumonia. Am J Respir Crit Care
Med. 2005;171:388–416.
2. Rybak M, Lomaestro B, Rotschafer JC, et al. Therapeutic
monitoring of vancomycin in adult patients: a consensus
review of the American Society of Health-System Pharma-
cists, the Infectious Diseases Society of America, and the
Society of Infectious Diseases Pharmacists. Am J Health Syst
Pharm. 2009;66:82–98.
3. Liu C, Bayer A, Cosgrove SE, et al. Clinical practice
guidelines by the Infectious Diseases Society of America for
the treatment of methicillin-resistant Staphylococcus au-
reus infections in adults and children. Clin Infect Dis.
2011;52:e18–e55.
4. Cantu TG, Yamanaka-Yuen NA, Lietman PS. Serum vanco-
mycin concentrations: reappraisal of their clinical value.
Clin Infect Dis. 1994;18:533–543.
5. Cunha BA, Mohan SS, Hamid N, et al. Cost-ineffectiveness
of serum vancomycin levels. Eur J Clin Microbiol Infect Dis.
2007;26:509–511.
6. Sorrell TC, Collignon PJ. A prospective study of adverse
reactions associated with vancomycin therapy. J Antimicrob
Chemother. 1985;16:235–241.
7. Wong-Beringer A, Joo J, Tse E, Beringer P. Vancomycin-
associated nephrotoxicity: a critical appraisal of risk with
high-dose therapy. Int J Antimicrob Agents. 2010;37:95–101.
8. Bosso JA, Nappi J, Rudisill C, et al. The relationship
betweenvancomycintroughconcentrationsandnephrotox-
icty: a prospective multicenter trial. Antimicrob Agents Che-
mother. 2011;55:5475–5479.
9. Jeffres MN, Isakow W, Doherty JA, et al. A retrospective
analysis of possible renal toxicity associated with vancomy-
cin in patients with health care-associated methicillin-
resistant Staphylococcus aureus pneumonia. Clin Ther.
2007;29:1107–1115.
10. Hermsen ED, Hanson M, Sankaranarayanan J, et al.
Clinical outcomes and nephrotoxicity associated with
vancomycin trough concentrations during treatment of
deep-seated infections. Expert Opin Drug Saf. 2010;9:9–14.
11. McKamy S, Hernandez E, Jahng M, et al. Incidence and risk
factorsinfluencingthedevelopmentofvancomycinnephro-
toxicity in children. J Pediatr. 2011;158:422–426.
12. Nguyen M, Wong J, Lee C, et al. Nephrotoxicity associated
with high dose vs standard dose vancomycin therapy.
Presented at: 47th Interscience Conference on Antimicro-
Clinical Therapeutics
156 Volume 34 Number 1
9. bialAgentsandChemotherapy;Sep-
tember 17–20, 2007; Chicago, Ill.
Abstract K-1096.
13. Prabaker K, Tran T, Pratummas T, et
al. Association of vancomycin
trough levels with nephrotoxicity.
Abstract presented at: 47th Annual
Meeting of Infectious Diseases Soci-
ety of America (IDSA); October 29
to November 1, 2009; Philadelphia,
Penn. Abstract 192.
14. Lodise TP, Lomaestro B, Graves J,
Drusano GL. Larger vancomycin
doses (at least four grams per day)
are associated with an increased
incidence of nephrotoxicity. Antimi-
crob Agents Chemother. 2008;52:
1330–1336.
15. Mangino JE, Peyrani P, Ford KD, et
al. Development and implementa-
tion of a performance improvement
project in adult intensive care units:
overview of the Improving Medicine
Through Pathway Assessment of
Critical Therapy in Hospital-Ac-
quired Pneumonia (IMPACT-HAP)
study. Crit Care. 2011;15:R38.
16. Knaus WA, Draper EA, Wagner DP,
Zimmerman JE. APACHE II: a sever-
ity of disease classification system.
Crit Care Med. 1985;13:818–829.
17. Lodise TP, Patel N, Lomaestro BM,
et al. Relationship between initial
vancomycinconcentration-timepro-
fileandnephrotoxicityamonghospi-
talized patients. Clin Infect Dis. 2009;
49:507–514.
18. Field A, Miles J. Discovering Statistics
Using SAS. Thousand Oaks, Calif:
Sage Publications Ltd; 2010.
19. Munro BH. Statistical Methods for
Health Care Research. Baltimore, Md:
LippincottWilliams&Wilkins;2004.
20. Vieira JM, Jr, Castro I, Curvello-Neto
A, et al. Effect of acute kidney injury
onweaningfrommechanicalventila-
tion in critically ill patients. Crit Care
Med. 2007;35:184–191.
21. Macedo E, Castro I, Yu L, et al.
Impact of mild acute kidney injury
(AKI) on outcome after open repair
of aortic aneurysms. Ren Fail. 2008;
30:287–296.
22. Chertow GM, Burdick E, Honour M,
et al. Acute kidney injury, mortality,
length of stay, and costs in hospital-
ized patients. J Am Soc Nephrol. 2005;
16:3365–3370.
23. Hidayat LK, Hsu DI, Quist R, et al.
High-dose vancomycin therapy for
methicillin-resistantStaphylococcusau-
reus infections: efficacy and toxicity.
Arch Intern Med. 2006;166:2138–2144.
24. Ingram PR, Lye DC, Tambyah PA, et
al. Risk factors for nephrotoxicity
associated with continuous vanco-
mycin infusion in outpatient paren-
teral antibiotic therapy. J Antimicrob
Chemother. 2008;62:168–171.
25. Matsko SM, Johns ST, Moise PA.
Impact of vancomycin (V) exposure
and other clinical characteristics on
nephrotoxicity.Abstractpresentedat:
48thAnnualMeetingoftheInfectious
Diseases Society of America; October
21–24, 2010; Vancouver, British Co-
lumbia,Canada.Abstract972.
26. Jeffres MN, Isakow W, Doherty JA,
et al. Predictors of mortality for
methicillin-resistant Staphylococ-
cus aureus health-care-associated
pneumonia: specific evaluation of
vancomycin pharmacokinetic indi-
ces. Chest. 2006;130:947–955.
27. Kett DH, Cano E, Quartin AA, et al.
Implementation of guidelines for
management of possible multidrug-
resistant pneumonia in intensive
care: an observational, multicentre
cohort study. Lancet Infect Dis. 2011;
11:181–189.
28. Venkataraman R, Kellum JA. Defin-
ing acute renal failure: the RIFLE
criteria.JIntensiveCareMed.2007;22:
187–193.
29. Mehta RL, Kellum JA, Shah SV, et al.
Acute Kidney Injury Network: report
of an initiative to improve outcomes
in acute kidney injury. Crit Care.
2007;11:R31.
Address correspondence to: Daniel H. Kett, MD, Division of Pulmonary
and Critical Care Medicine, the Miller School of Medicine at the University
of Miami, Jackson Memorial Hospital, 1611 NW 12th Avenue, C455A,
Miami, FL 33156. E-mail: dkett@med.miami.edu
E.L. Cano et al.
January 2012 157
![Incidence of Nephrotoxicity and Association With Vancomycin Use
in Intensive Care Unit Patients With Pneumonia: Retrospective
Analysis of the IMPACT-HAP Database
Ennie L. Cano, PharmD1
; Nadia Z. Haque, PharmD2
; Verna L. Welch, PhD, MPH3
;
Cynthia M. Cely, MD1
; Paula Peyrani, MD4
; Ernesto G. Scerpella, MD3
;
Kimbal D. Ford, PharmD3
; Marcus J. Zervos, MD5
; Julio A. Ramirez, MD4
; and
Daniel H. Kett, MD1
; on behalf of The Improving Medicine through Pathway Assessment
of Critical Therapy of Hospital-Acquired Pneumonia (IMPACT-HAP) Study Group*
1
Division of Pulmonary and Critical Care Medicine, University of Miami Miller School of Medicine and
Jackson Memorial Hospital, Miami, Florida; 2
Division of Infectious Diseases, Henry Ford Health System,
Detroit, Michigan; 3
Infectious Diseases, Specialty Care Medicines Development Group, Pfizer Inc,
Collegeville, Pennsylvania; 4
Division of Infectious Diseases, University of Louisville, Louisville, Kentucky;
and 5
Division of Infectious Diseases, Henry Ford Health System, Wayne State University School of
Medicine, Detroit, Michigan
ABSTRACT
Background: The 2005 guidelines from the Ameri-
can Thoracic Society and the Infectious Diseases Soci-
ety of America recommend vancomycin trough levels
of 15 to 20 mg/L for the therapy of hospital-acquired
(HAP), ventilator-associated (VAP), and health care–
associated (HCAP) pneumonia.
Objective: The goal of this article was to report the
incidence of nephrotoxicity and associated risk factors
in intensive care unit patients who received vancomy-
cin for the treatment of HAP, VAP, and HCAP.
Methods: This was a retrospective analysis of data
from a multicenter, observational study of pneumonia
patients. Antibiotic-associated nephrotoxicity was de-
fined as either an increase in serum creatinine Ն0.5
mg/dL or 50% above baseline, from initiation of van-
comycin to 72 hours after completion of therapy. Uni-
variate and multivariate logistic regression analyses
were performed to identify risk factors for develop-
ment of renal dysfunction.
Results: Of the 449 patients in the database, 240
received at least one dose of vancomycin and 188 had
sufficient data for analysis. In these 188 patients, 63%
were male. Mean (SD) age was 58.5 (17.2) years, and
the mean Acute Physiology and Chronic Health Eval-
uation II score was 19.4 (6.4). Nephrotoxicity oc-
curred in 29 of 188 (15.4%) vancomycin-treated pa-
tients. In multivariate analysis, initial vancomycin
trough levels Ն15 mg/L (odds ratio [OR], 5.2 [95% CI,
1.9–13.9]; P ϭ 0.001), concomitant aminoglycoside
use (OR, 2.67 [95% CI, 1.09–6.54]; P ϭ 0.03), and
duration of vancomycin therapy (OR for each addi-
tional treatment day, 1.12 [95% CI, 1.02–1.23]; P ϭ
0.02) were independently associated with nephrotox-
icity. The incidence of nephrotoxicity increased as a
function of the initial vancomycin trough level, rising
from 7% at a trough Ͻ10 mg/L to 34% at Ͼ20 mg/L
(P ϭ 0.001). The mean time to nephrotoxicity de-
creased from 8.8 days at vancomycin trough levels
Ͻ15 mg/L to 7.4 days at Ͼ20 mg/L (Kaplan-Meier
analysis, P ϭ 0.0003).
Conclusions: Nephrotoxicity may be common
among intensive care unit patients with pneumonia
treated with broad-spectrum antibiotic therapy that in-
cludes vancomycin. The finding that an initial vanco-
mycin trough level Ն15 mg/L may be an independent
This study was presented as a poster at the 48th Annual Interscience
Conference on Antimicrobial Agents and Chemotherapy and Infectious
Diseases Society of America 46th Annual Meeting, October 25–28,
2008, Washington, DC. It was also presented as a poster at the Infec-
tious Diseases Society of America 47th Annual Meeting, October 29 to
November 1, 2009, Philadelphia, Pennsylvania.
*The IMPACT-HAP Investigators are listed in the Acknowledg-
ments.
Accepted for publication December 15, 2011.
doi:10.1016/j.clinthera.2011.12.013
0149-2918/$ - see front matter
© 2012 Published by Elsevier HS Journals, Inc.
Clinical Therapeutics/Volume 34, Number 1, 2012
January 2012 149](https://image.slidesharecdn.com/2012vancomicinauci2012-170408150050/85/2012-vancomicina-uci2012-1-320.jpg)
