Kavitha K. Prabaker, MD

Methicillin‐resistant Staphylococcus aureus (MRSA) is responsible for an increasing number of invasive infections and, in the United States, may now be responsible for more deaths than disease associated with human immunodeficiency virus (HIV).1, 2 Vancomycin remains the drug of choice for invasive MRSA disease; from 1984 to 1996, its use in the United States escalated 6‐fold.3 With increased use of vancomycin, MRSA strains with partial and full resistance to vancomycin have emerged. Since 1997, S. aureus with intermediate susceptibility to vancomycin (VISA) as well as heteroresistance to vancomycin (hVISA) have been described.46 Several centers have also noted a slow rise in minimum inhibitory concentration (MIC) among clinical MRSA isolates (MIC creep).7 Low vancomycin trough levels have been implicated in the emergence of hVISA, and several studies have demonstrated a higher rate of vancomycin treatment failure, longer duration of fever, and prolonged hospitalization with hVISA and strains with elevated MIC compared to vancomycin‐susceptible MRSA.812 Until recently, vancomycin was frequently dosed to target trough levels <10 mg/L. The above concerns, combined with pharmacodynamic data suggesting that maintaining a ratio of vancomycin area under the curve to minimum inhibitory concentration (AUC/MIC) 400 may be associated with improved clinical outcome,13 have prompted an expert consensus to recommend targeting higher vancomycin trough levels (typically 15‐20 mg/L) for invasive MRSA infections and general avoidance of trough levels <10 mg/L.14

The effect of higher trough levels on kidney function remains poorly understood, as does the mechanism of vancomycin‐induced renal injury itself, though animal studies demonstrate oxidative damage to renal tubules with high doses of vancomycin.15, 16 In previous studies, the incidence of vancomycin nephrotoxicity with lower troughs has been reported to range from 0% to 19% with vancomycin alone, increasing up to 35% with concomitant aminoglycoside therapy.1724 Limited studies have been done to assess the risk of nephrotoxicity at higher trough levels. Lodise and colleagues identified high‐dose vancomycin (>4 gm per day) as an independent risk factor for nephrotoxicity, when compared to administration of <4 gm of vancomycin per day or use of linezolid, and showed greater risk of nephrotoxicity with increasing vancomycin trough levels within the first 96 hours of vancomycin administration.25, 26 Hidayat et al. demonstrated, in a prospective cohort analysis, that patients with mean trough levels 15 mg/L had a significantly increased incidence of nephrotoxicity. In that study, patients who developed nephrotoxicity were more likely to receive other nephrotoxic agents, and troughs collected before or after nephrotoxicity onset were not distinguished.9 This is an important distinction, as vancomycin is frequently continued with dose adjustment even after nephrotoxicity occurs, with the nephrotoxicity resulting in subsequent higher troughs. Jeffres et al. demonstrated that maximum vancomycin trough 15 mg/L was associated with nephrotoxicity in patients with healthcare‐associated MRSA pneumonia; this study was retrospective and focused on a particularly ill patient population.27 Pritchard et al. also retrospectively reviewed 2493 courses of vancomycin at their institution, from 2003 to 2007, and found a significant relationship between vancomycin trough 14 mg/L and nephrotoxicity. The presence of comorbid disease states and concomitant nephrotoxins was determined in a subset of 130 courses in 2007; increasing vancomycin trough was associated with nephrotoxicity in multivariable analysis.28 However, it is not clear whether troughs collected before or after nephrotoxicity onset were distinguished in this study. At least 6 other retrospective studies involving small sample size or published in abstract form have widely different results in relating high vancomycin trough or aggressive vancomycin dosing strategies to nephrotoxicity.2934

The purpose of our study was to evaluate the association between development of nephrotoxicity and trough levels obtained during vancomycin therapy at a large veterans' hospital, while accounting for other potential nephrotoxins, and to evaluate the temporal association between elevated vancomycin troughs and nephrotoxicity. We chose to focus on nephrotoxicity that occurred on, or after, 5 days of vancomycin therapy in order to reduce other confounding factors of nephrotoxicity, since short durations of vancomycin frequently represent use in surgical prophylaxis or empirical therapy for hemodynamically unstable patients at high risk for renal injury.

Patients and Methods

Results

DISCUSSION

Over the past 5 years, many institutions have adopted higher dosing guidelines for vancomycin, based on pharmacokinetic concerns related to its performance in the treatment of invasive staphylococcal disease. The data on nephrotoxicity at these higher troughs are limited. Previous studies that address the relationship between higher vancomycin troughs and nephrotoxicity suffer from small sample size29, 33; do not address reversibility of nephrotoxicity9, 26, 2931, 33; may not account for the temporal relationship between the development of nephrotoxicity and high trough levels,9, 2831 or examine patient populations at relatively high27 or low30 risk for renal injury apart from receipt of vancomycin. A recent expert consensus statement identified these factors as limiting the strength of evidence for a direct causal relationship between elevated vancomycin troughs and nephrotoxicity.14 A recent review by Hazlewood et al. concluded that the incidence of nephrotoxicity remains low in patients without preexisting renal disease and those not receiving concomitant nephrotoxins.35 The aim of our study was to identify whether or not there was a correlation between high‐dose vancomycin and nephrotoxicity, while accounting for their temporal relationship, concomitant nephrotoxin use, and reversibility. In particular, we chose to focus on nephrotoxicity occurring after at least 5 days of vancomycin therapy in order to reduce confounding by other possible sources of renal injury that may have affected the decision to initially prescribe vancomycin, an approach advocated by a recent review.36 While we noted that mean and maximum vancomycin troughs were significantly higher in 2007 than 2005‐2006, incidence of nephrotoxicity was stable between the 2 time periods, with the higher rate of intravenous contrast dye in 2007 balanced in part by less aminoglycoside use. Overall, higher trough levels were not necessarily accompanied by a significant increase in nephrotoxicity, though there was a nonsignificant trend toward more nephrotoxic patients having maximum trough 15 mg/L.

The only clinical factor that was significantly associated with nephrotoxicity in multivariate analysis was receipt of intravenous contrast dye. Of the 44 patients who received intravenous contrast dye, 10 (22.7%) experienced nephrotoxicity. Interestingly, in animal studies, both intravenous contrast dye37, 38 and high‐dose vancomycin15, 16 have been demonstrated to promote free radical formation within renal tissue, which is hypothesized to cause tubular damage primarily through vascular endothelial dysfunction, vasoconstriction, and subsequent reperfusion injury. N‐acetylcysteine is frequently administered to patients about to receive intravenous contrast dye (although its benefit remains controversial37, 39); N‐acetylcysteine has also been shown in an animal model to attenuate vancomycin‐induced renal injury.40

Receipt of concomitant aminoglycosides was not significantly associated with nephrotoxicity, in contrast with previous studies. One meta‐analysis of 8 studies revealed found that the incidence of nephrotoxicity associated with combination vancomycin and aminoglycosides was 13.3% greater than with vancomycin alone (P < 0.01) and 4.3% greater than therapy with an aminoglycoside alone (P < 0.05)20; another analysis of safety data of the clinical trial comparing daptomycin to comparator therapy including initial low‐dose gentamicin therapy in the treatment of S. aureus bacteremia found renal adverse events in 10 of 53 (19%) patients receiving vancomycin and gentamicin, compared to 8 of 120 (7%) patients receiving daptomycin.41 While our findings that show no clear relationship between concomitant vancomycin and aminoglycoside use and nephrotoxicity may have been due to the relatively small number of patients in our study who received aminoglycosides, it is worth noting that more patients in our study received aminoglycosides than intravenous contrast dye (66 vs 44 patients). The 77.8% overall resolution of nephrotoxicity observed in our study is similar to that reported by Farber and Moellering in 198319 and to that reported more recently with high‐dose vancomycin by Jeffres et al. and Teng et al.27, 34

Although we attempted to account for as many confounders as possible, the retrospective nature of our study prevents us from making definitive statements regarding the role of vancomycin trough levels and nephrotoxicity. In particular, we are unable to comment on any potential role vancomycin may have on nephrotoxicity within 5 days of its start or on patients with a baseline serum creatinine >2. Other significant limitations include our small proportion of female patients, and that we were not able to calculate severity of illness or determine the presence of congestive heart failure. Also, we may be dosing vancomycin less aggressively than other centers, and thus may have reduced power in determining whether higher troughs, particularly those 20 mg/L, are associated with nephrotoxicity; identification of more patients with higher troughs and a larger overall sample size may have yielded different results. Even in the 2007 group, a significant number of patients with cellulitis, UTI, and uncomplicated bacteremia had target troughs of 8‐12 mg/L. However, taken together, our findings do not support a definite relationship between vancomycin troughs and development of nephrotoxicity, and that when it does occur, it is largely reversible. Further prospective research is needed to evaluate the effects of aggressive vancomycin dosing regimens on nephrotoxicity, particularly those regimens that include large loading doses. Trials of antioxidative agents in patients receiving aggressive dosing regimens of vancomycin who require radiology studies involving intravenous contrast dye may be indicated as well.

Methicillin‐resistant Staphylococcus aureus (MRSA) is responsible for an increasing number of invasive infections and, in the United States, may now be responsible for more deaths than disease associated with human immunodeficiency virus (HIV).1, 2 Vancomycin remains the drug of choice for invasive MRSA disease; from 1984 to 1996, its use in the United States escalated 6‐fold.3 With increased use of vancomycin, MRSA strains with partial and full resistance to vancomycin have emerged. Since 1997, S. aureus with intermediate susceptibility to vancomycin (VISA) as well as heteroresistance to vancomycin (hVISA) have been described.46 Several centers have also noted a slow rise in minimum inhibitory concentration (MIC) among clinical MRSA isolates (MIC creep).7 Low vancomycin trough levels have been implicated in the emergence of hVISA, and several studies have demonstrated a higher rate of vancomycin treatment failure, longer duration of fever, and prolonged hospitalization with hVISA and strains with elevated MIC compared to vancomycin‐susceptible MRSA.812 Until recently, vancomycin was frequently dosed to target trough levels <10 mg/L. The above concerns, combined with pharmacodynamic data suggesting that maintaining a ratio of vancomycin area under the curve to minimum inhibitory concentration (AUC/MIC) 400 may be associated with improved clinical outcome,13 have prompted an expert consensus to recommend targeting higher vancomycin trough levels (typically 15‐20 mg/L) for invasive MRSA infections and general avoidance of trough levels <10 mg/L.14

The effect of higher trough levels on kidney function remains poorly understood, as does the mechanism of vancomycin‐induced renal injury itself, though animal studies demonstrate oxidative damage to renal tubules with high doses of vancomycin.15, 16 In previous studies, the incidence of vancomycin nephrotoxicity with lower troughs has been reported to range from 0% to 19% with vancomycin alone, increasing up to 35% with concomitant aminoglycoside therapy.1724 Limited studies have been done to assess the risk of nephrotoxicity at higher trough levels. Lodise and colleagues identified high‐dose vancomycin (>4 gm per day) as an independent risk factor for nephrotoxicity, when compared to administration of <4 gm of vancomycin per day or use of linezolid, and showed greater risk of nephrotoxicity with increasing vancomycin trough levels within the first 96 hours of vancomycin administration.25, 26 Hidayat et al. demonstrated, in a prospective cohort analysis, that patients with mean trough levels 15 mg/L had a significantly increased incidence of nephrotoxicity. In that study, patients who developed nephrotoxicity were more likely to receive other nephrotoxic agents, and troughs collected before or after nephrotoxicity onset were not distinguished.9 This is an important distinction, as vancomycin is frequently continued with dose adjustment even after nephrotoxicity occurs, with the nephrotoxicity resulting in subsequent higher troughs. Jeffres et al. demonstrated that maximum vancomycin trough 15 mg/L was associated with nephrotoxicity in patients with healthcare‐associated MRSA pneumonia; this study was retrospective and focused on a particularly ill patient population.27 Pritchard et al. also retrospectively reviewed 2493 courses of vancomycin at their institution, from 2003 to 2007, and found a significant relationship between vancomycin trough 14 mg/L and nephrotoxicity. The presence of comorbid disease states and concomitant nephrotoxins was determined in a subset of 130 courses in 2007; increasing vancomycin trough was associated with nephrotoxicity in multivariable analysis.28 However, it is not clear whether troughs collected before or after nephrotoxicity onset were distinguished in this study. At least 6 other retrospective studies involving small sample size or published in abstract form have widely different results in relating high vancomycin trough or aggressive vancomycin dosing strategies to nephrotoxicity.2934

The purpose of our study was to evaluate the association between development of nephrotoxicity and trough levels obtained during vancomycin therapy at a large veterans' hospital, while accounting for other potential nephrotoxins, and to evaluate the temporal association between elevated vancomycin troughs and nephrotoxicity. We chose to focus on nephrotoxicity that occurred on, or after, 5 days of vancomycin therapy in order to reduce other confounding factors of nephrotoxicity, since short durations of vancomycin frequently represent use in surgical prophylaxis or empirical therapy for hemodynamically unstable patients at high risk for renal injury.

Patients and Methods

Results

DISCUSSION

Over the past 5 years, many institutions have adopted higher dosing guidelines for vancomycin, based on pharmacokinetic concerns related to its performance in the treatment of invasive staphylococcal disease. The data on nephrotoxicity at these higher troughs are limited. Previous studies that address the relationship between higher vancomycin troughs and nephrotoxicity suffer from small sample size29, 33; do not address reversibility of nephrotoxicity9, 26, 2931, 33; may not account for the temporal relationship between the development of nephrotoxicity and high trough levels,9, 2831 or examine patient populations at relatively high27 or low30 risk for renal injury apart from receipt of vancomycin. A recent expert consensus statement identified these factors as limiting the strength of evidence for a direct causal relationship between elevated vancomycin troughs and nephrotoxicity.14 A recent review by Hazlewood et al. concluded that the incidence of nephrotoxicity remains low in patients without preexisting renal disease and those not receiving concomitant nephrotoxins.35 The aim of our study was to identify whether or not there was a correlation between high‐dose vancomycin and nephrotoxicity, while accounting for their temporal relationship, concomitant nephrotoxin use, and reversibility. In particular, we chose to focus on nephrotoxicity occurring after at least 5 days of vancomycin therapy in order to reduce confounding by other possible sources of renal injury that may have affected the decision to initially prescribe vancomycin, an approach advocated by a recent review.36 While we noted that mean and maximum vancomycin troughs were significantly higher in 2007 than 2005‐2006, incidence of nephrotoxicity was stable between the 2 time periods, with the higher rate of intravenous contrast dye in 2007 balanced in part by less aminoglycoside use. Overall, higher trough levels were not necessarily accompanied by a significant increase in nephrotoxicity, though there was a nonsignificant trend toward more nephrotoxic patients having maximum trough 15 mg/L.

The only clinical factor that was significantly associated with nephrotoxicity in multivariate analysis was receipt of intravenous contrast dye. Of the 44 patients who received intravenous contrast dye, 10 (22.7%) experienced nephrotoxicity. Interestingly, in animal studies, both intravenous contrast dye37, 38 and high‐dose vancomycin15, 16 have been demonstrated to promote free radical formation within renal tissue, which is hypothesized to cause tubular damage primarily through vascular endothelial dysfunction, vasoconstriction, and subsequent reperfusion injury. N‐acetylcysteine is frequently administered to patients about to receive intravenous contrast dye (although its benefit remains controversial37, 39); N‐acetylcysteine has also been shown in an animal model to attenuate vancomycin‐induced renal injury.40

Receipt of concomitant aminoglycosides was not significantly associated with nephrotoxicity, in contrast with previous studies. One meta‐analysis of 8 studies revealed found that the incidence of nephrotoxicity associated with combination vancomycin and aminoglycosides was 13.3% greater than with vancomycin alone (P < 0.01) and 4.3% greater than therapy with an aminoglycoside alone (P < 0.05)20; another analysis of safety data of the clinical trial comparing daptomycin to comparator therapy including initial low‐dose gentamicin therapy in the treatment of S. aureus bacteremia found renal adverse events in 10 of 53 (19%) patients receiving vancomycin and gentamicin, compared to 8 of 120 (7%) patients receiving daptomycin.41 While our findings that show no clear relationship between concomitant vancomycin and aminoglycoside use and nephrotoxicity may have been due to the relatively small number of patients in our study who received aminoglycosides, it is worth noting that more patients in our study received aminoglycosides than intravenous contrast dye (66 vs 44 patients). The 77.8% overall resolution of nephrotoxicity observed in our study is similar to that reported by Farber and Moellering in 198319 and to that reported more recently with high‐dose vancomycin by Jeffres et al. and Teng et al.27, 34

Although we attempted to account for as many confounders as possible, the retrospective nature of our study prevents us from making definitive statements regarding the role of vancomycin trough levels and nephrotoxicity. In particular, we are unable to comment on any potential role vancomycin may have on nephrotoxicity within 5 days of its start or on patients with a baseline serum creatinine >2. Other significant limitations include our small proportion of female patients, and that we were not able to calculate severity of illness or determine the presence of congestive heart failure. Also, we may be dosing vancomycin less aggressively than other centers, and thus may have reduced power in determining whether higher troughs, particularly those 20 mg/L, are associated with nephrotoxicity; identification of more patients with higher troughs and a larger overall sample size may have yielded different results. Even in the 2007 group, a significant number of patients with cellulitis, UTI, and uncomplicated bacteremia had target troughs of 8‐12 mg/L. However, taken together, our findings do not support a definite relationship between vancomycin troughs and development of nephrotoxicity, and that when it does occur, it is largely reversible. Further prospective research is needed to evaluate the effects of aggressive vancomycin dosing regimens on nephrotoxicity, particularly those regimens that include large loading doses. Trials of antioxidative agents in patients receiving aggressive dosing regimens of vancomycin who require radiology studies involving intravenous contrast dye may be indicated as well.

Methicillin‐resistant Staphylococcus aureus (MRSA) is responsible for an increasing number of invasive infections and, in the United States, may now be responsible for more deaths than disease associated with human immunodeficiency virus (HIV).1, 2 Vancomycin remains the drug of choice for invasive MRSA disease; from 1984 to 1996, its use in the United States escalated 6‐fold.3 With increased use of vancomycin, MRSA strains with partial and full resistance to vancomycin have emerged. Since 1997, S. aureus with intermediate susceptibility to vancomycin (VISA) as well as heteroresistance to vancomycin (hVISA) have been described.46 Several centers have also noted a slow rise in minimum inhibitory concentration (MIC) among clinical MRSA isolates (MIC creep).7 Low vancomycin trough levels have been implicated in the emergence of hVISA, and several studies have demonstrated a higher rate of vancomycin treatment failure, longer duration of fever, and prolonged hospitalization with hVISA and strains with elevated MIC compared to vancomycin‐susceptible MRSA.812 Until recently, vancomycin was frequently dosed to target trough levels <10 mg/L. The above concerns, combined with pharmacodynamic data suggesting that maintaining a ratio of vancomycin area under the curve to minimum inhibitory concentration (AUC/MIC) 400 may be associated with improved clinical outcome,13 have prompted an expert consensus to recommend targeting higher vancomycin trough levels (typically 15‐20 mg/L) for invasive MRSA infections and general avoidance of trough levels <10 mg/L.14

The effect of higher trough levels on kidney function remains poorly understood, as does the mechanism of vancomycin‐induced renal injury itself, though animal studies demonstrate oxidative damage to renal tubules with high doses of vancomycin.15, 16 In previous studies, the incidence of vancomycin nephrotoxicity with lower troughs has been reported to range from 0% to 19% with vancomycin alone, increasing up to 35% with concomitant aminoglycoside therapy.1724 Limited studies have been done to assess the risk of nephrotoxicity at higher trough levels. Lodise and colleagues identified high‐dose vancomycin (>4 gm per day) as an independent risk factor for nephrotoxicity, when compared to administration of <4 gm of vancomycin per day or use of linezolid, and showed greater risk of nephrotoxicity with increasing vancomycin trough levels within the first 96 hours of vancomycin administration.25, 26 Hidayat et al. demonstrated, in a prospective cohort analysis, that patients with mean trough levels 15 mg/L had a significantly increased incidence of nephrotoxicity. In that study, patients who developed nephrotoxicity were more likely to receive other nephrotoxic agents, and troughs collected before or after nephrotoxicity onset were not distinguished.9 This is an important distinction, as vancomycin is frequently continued with dose adjustment even after nephrotoxicity occurs, with the nephrotoxicity resulting in subsequent higher troughs. Jeffres et al. demonstrated that maximum vancomycin trough 15 mg/L was associated with nephrotoxicity in patients with healthcare‐associated MRSA pneumonia; this study was retrospective and focused on a particularly ill patient population.27 Pritchard et al. also retrospectively reviewed 2493 courses of vancomycin at their institution, from 2003 to 2007, and found a significant relationship between vancomycin trough 14 mg/L and nephrotoxicity. The presence of comorbid disease states and concomitant nephrotoxins was determined in a subset of 130 courses in 2007; increasing vancomycin trough was associated with nephrotoxicity in multivariable analysis.28 However, it is not clear whether troughs collected before or after nephrotoxicity onset were distinguished in this study. At least 6 other retrospective studies involving small sample size or published in abstract form have widely different results in relating high vancomycin trough or aggressive vancomycin dosing strategies to nephrotoxicity.2934

The purpose of our study was to evaluate the association between development of nephrotoxicity and trough levels obtained during vancomycin therapy at a large veterans' hospital, while accounting for other potential nephrotoxins, and to evaluate the temporal association between elevated vancomycin troughs and nephrotoxicity. We chose to focus on nephrotoxicity that occurred on, or after, 5 days of vancomycin therapy in order to reduce other confounding factors of nephrotoxicity, since short durations of vancomycin frequently represent use in surgical prophylaxis or empirical therapy for hemodynamically unstable patients at high risk for renal injury.

Patients and Methods

Results

DISCUSSION

Over the past 5 years, many institutions have adopted higher dosing guidelines for vancomycin, based on pharmacokinetic concerns related to its performance in the treatment of invasive staphylococcal disease. The data on nephrotoxicity at these higher troughs are limited. Previous studies that address the relationship between higher vancomycin troughs and nephrotoxicity suffer from small sample size29, 33; do not address reversibility of nephrotoxicity9, 26, 2931, 33; may not account for the temporal relationship between the development of nephrotoxicity and high trough levels,9, 2831 or examine patient populations at relatively high27 or low30 risk for renal injury apart from receipt of vancomycin. A recent expert consensus statement identified these factors as limiting the strength of evidence for a direct causal relationship between elevated vancomycin troughs and nephrotoxicity.14 A recent review by Hazlewood et al. concluded that the incidence of nephrotoxicity remains low in patients without preexisting renal disease and those not receiving concomitant nephrotoxins.35 The aim of our study was to identify whether or not there was a correlation between high‐dose vancomycin and nephrotoxicity, while accounting for their temporal relationship, concomitant nephrotoxin use, and reversibility. In particular, we chose to focus on nephrotoxicity occurring after at least 5 days of vancomycin therapy in order to reduce confounding by other possible sources of renal injury that may have affected the decision to initially prescribe vancomycin, an approach advocated by a recent review.36 While we noted that mean and maximum vancomycin troughs were significantly higher in 2007 than 2005‐2006, incidence of nephrotoxicity was stable between the 2 time periods, with the higher rate of intravenous contrast dye in 2007 balanced in part by less aminoglycoside use. Overall, higher trough levels were not necessarily accompanied by a significant increase in nephrotoxicity, though there was a nonsignificant trend toward more nephrotoxic patients having maximum trough 15 mg/L.

The only clinical factor that was significantly associated with nephrotoxicity in multivariate analysis was receipt of intravenous contrast dye. Of the 44 patients who received intravenous contrast dye, 10 (22.7%) experienced nephrotoxicity. Interestingly, in animal studies, both intravenous contrast dye37, 38 and high‐dose vancomycin15, 16 have been demonstrated to promote free radical formation within renal tissue, which is hypothesized to cause tubular damage primarily through vascular endothelial dysfunction, vasoconstriction, and subsequent reperfusion injury. N‐acetylcysteine is frequently administered to patients about to receive intravenous contrast dye (although its benefit remains controversial37, 39); N‐acetylcysteine has also been shown in an animal model to attenuate vancomycin‐induced renal injury.40

Receipt of concomitant aminoglycosides was not significantly associated with nephrotoxicity, in contrast with previous studies. One meta‐analysis of 8 studies revealed found that the incidence of nephrotoxicity associated with combination vancomycin and aminoglycosides was 13.3% greater than with vancomycin alone (P < 0.01) and 4.3% greater than therapy with an aminoglycoside alone (P < 0.05)20; another analysis of safety data of the clinical trial comparing daptomycin to comparator therapy including initial low‐dose gentamicin therapy in the treatment of S. aureus bacteremia found renal adverse events in 10 of 53 (19%) patients receiving vancomycin and gentamicin, compared to 8 of 120 (7%) patients receiving daptomycin.41 While our findings that show no clear relationship between concomitant vancomycin and aminoglycoside use and nephrotoxicity may have been due to the relatively small number of patients in our study who received aminoglycosides, it is worth noting that more patients in our study received aminoglycosides than intravenous contrast dye (66 vs 44 patients). The 77.8% overall resolution of nephrotoxicity observed in our study is similar to that reported by Farber and Moellering in 198319 and to that reported more recently with high‐dose vancomycin by Jeffres et al. and Teng et al.27, 34

Although we attempted to account for as many confounders as possible, the retrospective nature of our study prevents us from making definitive statements regarding the role of vancomycin trough levels and nephrotoxicity. In particular, we are unable to comment on any potential role vancomycin may have on nephrotoxicity within 5 days of its start or on patients with a baseline serum creatinine >2. Other significant limitations include our small proportion of female patients, and that we were not able to calculate severity of illness or determine the presence of congestive heart failure. Also, we may be dosing vancomycin less aggressively than other centers, and thus may have reduced power in determining whether higher troughs, particularly those 20 mg/L, are associated with nephrotoxicity; identification of more patients with higher troughs and a larger overall sample size may have yielded different results. Even in the 2007 group, a significant number of patients with cellulitis, UTI, and uncomplicated bacteremia had target troughs of 8‐12 mg/L. However, taken together, our findings do not support a definite relationship between vancomycin troughs and development of nephrotoxicity, and that when it does occur, it is largely reversible. Further prospective research is needed to evaluate the effects of aggressive vancomycin dosing regimens on nephrotoxicity, particularly those regimens that include large loading doses. Trials of antioxidative agents in patients receiving aggressive dosing regimens of vancomycin who require radiology studies involving intravenous contrast dye may be indicated as well.