Title: Effect of Continuous Venovenous Hemofiltration Dose on the Achievement of Adequate Vancomycin Trough Concentrations

AAC Accepts, published online ahead of print on 17 September 2012 Antimicrob. Agents Chemother. doi:10.1128/aac.00459-12 Copyright 2012, American Society for Microbiology. All Rights Reserved. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Title: Effect of Continuous Venovenous Hemofiltration Dose on the Achievement of Adequate Vancomycin Trough Concentrations Running title: CVVH Dose and Vancomycin Trough Achievement Erin N. Frazee, PharmD# 1 Philip J. Kuper, PharmD 1 Garrett E. Schramm, PharmD 1 Scott L. Larson, PharmD 1 Kianoush B. Kashani, MD 2,3 Douglas R. Osmon, MD 4 Nelson Leung, MD 2 1 Hospital Pharmacy Services, Mayo Clinic, Rochester, MN 2 Division of Nephrology, Mayo Clinic, Rochester, MN 3 Division of Critical Care Medicine, Mayo Clinic, Rochester, MN 4 Division of Infectious Diseases, Mayo Clinic, Rochester, MN Institution at which work was performed: Mayo Clinic 200 First St SW Rochester, MN 55905 Corresponding author: Erin N. Frazee Email address: frazee.erin@mayo.edu

Page 1 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 Abstract The vancomycin dose necessary for the achievement of target serum trough concentrations during continuous venovenous hemofiltration (CVVH) remains to be elucidated. This was a retrospective cohort study of critically ill adults at a tertiary medical center on concurrent CVVH and vancomycin between 2006 and 2010 with a steady state vancomycin trough concentration. The 87 included patients were grouped according to low ( 30 ml/kg/h; N = 10) or high (> 30 ml/kg/h; N = 77) CVVH hemofiltration rate (HFR) for analysis. Vancomycin goal trough achievement occurred in only 32 (37%) patients. The primary endpoint of trough attainment significantly differed between HFR subgroups, 90% versus 30% in low and high HFR individuals, respectively (P < 0.001). Patients with subtherapeutic trough concentrations had a median (interquartile range) HFR of 40 (37-47) ml/kg/hr compared to 36 (30-39) ml/kg/hr in those who achieved the trough goal. Irrespective of goal trough, an inverse correlation existed between HFR and serum vancomycin concentration (R = - 0.423; P < 0.001). In the subgroup of 14 methicillin resistant Staphylococcus aureus (MRSA) patients, trough achievement was similar to the aggregate cohort (36%). Mortality at 28 days was unrelated to trough achievement in both the overall sample (P = 0.516) and in culture-positive MRSA patients (P = 0.396). Critically ill patients undergoing CVVH therapy may experience clinically significant reductions in goal vancomycin troughs. The results of this study justify prospective evaluations in this population to determine the optimal vancomycin dosing strategy for attainment of goal trough concentrations. 48

Page 2 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 Introduction Severe sepsis contributes to 10-20% of intensive care unit (ICU) admissions and is associated with an estimated 30% mortality rate (2, 24, 25). Sepsis-related sequelae precipitate nearly 50% of all acute kidney injury cases which often progress to require renal replacement therapy(1, 3, 31). In a 2003 multicenter practice survey, 86.2% of providers reported utilization of continuous renal replacement therapy (CRRT) as one of their management strategies for acute kidney injury(18). Provision of CRRT accomplishes steady correction of metabolic abnormalities and fluid balance without exacerbating unstable hemodynamics(17, 22). Use of continuous venovenous hemofiltration (CVVH), a convective clearance CRRT modality, has become the preferred modality likely due to the poor solute removal and arterial cannulation complications associated with arteriovenous circuits(17, 18, 20, 28). Hemofiltration rate impacts solute removal during CVVH (8). Early CVVH literature demonstrated an improvement in mortality with hemofiltration rates of 35-45 ml/kg/hr when compared to rates of 20 ml/kg/hr(22). The conflicting results of subsequent studies on the appropriate dose of CRRT have lead to varied application in clinical practice(4, 6, 19). Despite this variability, it is clear that an upward titration of hemofiltration rates occurred in the last decade(4, 6, 19, 22, 26). The importance of rate-dependent solute removal is magnified when the solute is an essential pharmacotherapeutic intervention. Evolution of hemofiltration rate practices may impact antimicrobial agent adequacy resulting in adverse clinical outcomes in the septic patient (27). The Surviving Sepsis guideline recommendations stress not only activity of anti-infective agents against all likely pathogens, but also pharmacokinetic/pharmacodynamic optimization to ensure maximal efficacy (10). Continuous venovenous hemofiltration is known to impact the clearance of vancomycin, an antimicrobial often recommended for empiric coverage in sepsis(5,

Page 3 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 10, 12, 14, 23). Published literature on vancomycin clearance during CVVH is limited to small sample sizes, limited patient populations, and lower hemofiltration rates than those currently used (5, 9, 12, 21). With the change in CVVH practices, adequacy of existing vancomycin dosing strategies has not been investigated. The purpose of this study was to determine the success rate of an institutional vancomycin dosing recommendation in achieving target trough concentrations in patients on CVVH. The investigators hypothesized that use of high hemofiltration rates would portend failure to achieve goal troughs more often than low hemofiltration rates. Materials and Methods Subject identification and data collection This retrospective cohort study examined adult ICU patients at Mayo Clinic in Rochester, Minnesota, who received concurrent CVVH and vancomycin between January 2006 and December 2010. The institutional review board approved this study and waived the need for informed consent. An existing CVVH database identified eligible patients who were then grouped according to low ( 30 ml/kg/hr) or high (> 30 ml/kg/hr) hemofiltration rate(6, 7, 19, 21, 22, 26, 28). The study included adults aged 18 years who received 15-20 mg/kg actual body weight (ABW) vancomycin as a single daily dose consistent with institutional recommendations(5, 12, 27). Goal vancomycin trough concentrations were in accordance with Infectious Diseases Society of America guideline recommendations, 10-15 mg/l or 15-20 mg/l, based on suspected or documented source(s) of infection(23). Trough concentrations were drawn no earlier than before the fourth consistent vancomycin dose or the third if given a loading dose to assure steady-state conditions(5, 11). Vancomycin concentration analysis utilized an Olympus AU680 (Olympus America, Inc., Melville, NY) with the Syva Emit 2000 Vancomycin Assay; Siemens

Page 4 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 Healthcare Diagnostics, Inc (Newark, DE). The assay has an analytical range between 5.0 50.0 mg/l and the between-run coefficient of variation was < 10% throughout the analytical range All CVVH treatments applied the Prismaflex System and the HF 1400 polyarylethersulfone filter (Gambro, Stockholm, Sweden). Blood flow rate was standard at 200 ml/min and anticoagulation achieved with sodium citrate. Prismasate (Gambro, Inc. Lakewood, CO) replaced the hemofiltration fluid, 50% pre-filter and 50% post-filter. The attending nephrologist prescribed an individualized hemofiltration rate for each patient. Excluded patients did not authorize their medical chart for review, received less than 85% of their prescribed hemofiltration rate, had CVVH held for more than eight consecutive hours during vancomycin therapy, required group crossover due to hemofiltration rate prescription changes, had 0.5 ml/kg/hour average urine output for six consecutive hours during the study, or received concurrent extracorporeal membrane oxygenation (10, 19, 22, 30). Outcome analysis included only the first ICU admission for each patient during the study timeframe. Demographic data [age, sex, ABW, ideal body weight, body mass index (BMI), baseline renal function, comorbid conditions], severity of illness scores [Acute Physiology and Chronic Health Evaluation II (APACHE II), Sequential Organ Failure Assessment (SOFA)], and laboratory parameters (serum albumin, blood urea nitrogen) were retrospectively collected for each patient in the study. Other gathered data included admitting diagnosis, primary ICU service, source of infection, renal parameters (daily CVVH hemofiltration rate, dialyzer used, fluid balance, urine output), weight-based vancomycin dose, and serum trough concentration. Endpoints The primary outcome measured the success rate of vancomycin goal trough achievement between low and high hemofiltration rate groups. Secondary outcomes included trough

Page 5 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 concentrations in the aggregate cohort, factors associated with hemofiltration rate prescription, and 28-day all-cause mortality. Data analysis No formal sample size calculation was able to be performed and all eligible patients who met inclusion/exclusion criteria were analyzed. Descriptive data were summarized by medians with interquartile ranges (IQR) and percentages. The Chi-square test (or Fisher s exact test as appropriate) studied categorical variables and the Wilcoxon rank sum test analyzed continuous variables. The Spearman rank correlation coefficient tested the association between two continuous variables. Patient-specific characteristics were imputed in a multivariate least squares regression model to analyze factors associated with continuous variables. Survival rates were estimated and compared between subgroups using the Kaplan-Meier method with the logrank test. All analyses were carried out using the JMP statistical software package (Version 8, SAS Institute Inc., Cary, NC). A P value less than 0.05 was considered statistically significant. Results Patients During the study period, 435 patients underwent evaluation for eligibility (Figure 1). The majority of the 348 excluded patients lacked an available steady-state trough concentration (N = 277; 80%) or received vancomycin at a dose different from the institutional recommendation (N = 38; 11%). The 87 included patients were a median (IQR) age of 62 (55-72) years and 64% male. The sample was predominately comprised of mixed-surgical ICU (43%) and medical ICU (26%) patients with critical illness reflected in APACHE II and SOFA scores of 25 (16-34) and 12 (7-15), respectively. Pulmonary (N = 44) and intra-abdominal (N = 19) sources of infection occurred most commonly. Few differences existed between groups at baseline (Table 1). Hemofiltration rates were 27 (24-30) ml/kg/hr and 39 (36-45) ml/kg/hr in the low (N = 10) and

Page 6 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 high (N = 77) rate groups, respectively (P < 0.0001). The low hemofiltration rate group had a significantly higher BMI than high hemofiltration rate patients (P < 0.001) and proportionally more intra-abdominal infections occurred in low rate individuals (P = 0.037). First vancomycin dose which met study definition was administered -0.1 (-0.5-1.6) days from CVVH initiation. Twenty two patients (25%) received a loading dose, all of which were in the high hemofiltration rate group. Loading doses were approximately 24.0 (20.3-26.4) mg/kg. Microbiologic confirmation of infection was present in 43 (49%) cases. Fourteen patients (16%) developed culture-positive methicillin resistant Staphylococcus aureus (MRSA) infections, of which 7 (50%) experienced an MRSA bacteremia. Severity of illness at ICU admission in this subgroup compared well to the aggregate cohort with median (IQR) APACHE II score of 25 (18-37) and SOFA score of 11 (7-13). Baseline BMI for these patients was 28 (23-34) kg/m 2 and they were prescribed a hemofiltration rate of 41 (37-49) ml/kg/hr. Endpoints Trough concentrations were obtained a median (IQR) of 3.0 (3.0-4.0) days after vancomycin therapy commenced. Overall goal trough attainment occurred in 32 (37%) patients. For the primary endpoint of trough achievement according to hemofiltration rate subgroup, low rate patients achieved goal trough concentrations 90% of the time while only 30% of high hemofiltration rate cases reached their goal (P < 0.001). Patients with subtherapeutic trough concentrations had hemofiltration rates of 40 (37-47) ml/kg/hr, in contrast to hemofiltration rates of 36 (30-39) ml/kg/hr in patients who reached goal trough levels (P < 0.001). Irrespective of trough goal, a significant inverse correlation exists between hemofiltration rate and vancomycin serum concentration (Figure 2; P < 0.001). Mortality at 28 days did not differ according to trough achievement (Figure 3; P = 0.516).

Page 7 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 In the subgroup of 14 patients with a documented MRSA infection, two (14%) patients were prescribed 30 ml/kg/hr hemofiltration rate, whereas 12 (86%) received > 30 ml/kg/hr. Nine MRSA patients (64%) had subtherapeutic trough concentrations. Both low hemofiltration rate patients were within goal range, whereas nine (75%) high hemofiltration rate patients manifested subtherapeutic troughs (P = 0.110). Similar to the aggregate sample, no difference in 28-day mortality was noted between trough achievers and non-achievers (P = 0.396). Nephrologists individualized CVVH hemofiltration rate prescriptions for each patient and variability in prescribing practices existed among practitioners. Hemofiltration rates in the aggregate cohort ranged from 12-87 ml/kg/hr. Prescribed hemofiltration rate (ml/kg/hr) was independently associated with ABW (P < 0.001), sex (P = 0.002), BMI (P = 0.006), and baseline serum creatinine (P = 0.03) in a multivariate analysis. Discussion In this retrospective cohort analysis of CVVH patients, vancomycin goal trough concentration achievement occurred significantly more often in individuals receiving low hemofiltration rates compared to patients receiving high hemofiltration rates. Increases in hemofiltration rates regardless of vancomycin goal range significantly correlated with reductions in trough concentrations. Similar to the aggregate cohort, in the subset of patients with culturepositive MRSA infections, the majority of patients in the high hemofiltration rate group demonstrated subtherapeutic trough concentrations. Hemofiltration rate prescribing practices varied among providers and factors shown to be independently associated with the prescribed rate (ml/kg/hr) included ABW, sex, BMI, and baseline serum creatinine These findings are clinically relevant due to the paucity of available vancomycin pharmacokinetic data during CVVH. Joy, et al. characterized the clearance of vancomycin during CVVH in eight non-critically ill end-stage renal disease patients. Each patient received a

Page 8 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 500 mg dose and CVVH clearance of vancomycin depended on hemofiltration rates when assessed at 500 ml/hr and 1000 ml/hr(12). Boereboom, et al. described vancomycin clearance during CVVH in two cases of septic shock and multiple organ dysfunction syndrome. Determination of vancomycin pharmacokinetics occurred after a single dose in one individual and on the sixth day of therapy in the other study subject. Apparent volumes of distribution (V d ) were 0.66 and 0.52 L/kg, respectively, and terminal half-lives were 15.4 and 20.3 hours for vancomycin. The CVVH clearance of vancomycin was 1400 ml/hr at a hemofiltration rate of 1500 ml/hr (19 and 20 ml/kg/hr) in the two patients(5). New evidence tested these findings in seven patients given a single 12.5-20 mg/kg vancomycin dose during CVVH at hemofiltration rates of 800-1200 ml/hr(9). The calculated sieving coefficient of 0.71 ± 0.13 approximated existing literature (0.70 ± 0.15 0.89 ± 0.03) (5, 9, 12). Chaijamorn, et al. identified vancomycin CVVH clearance of 730 ml/hr, lower than that documented in prior studies, and 50% of total clearance attributable to non-renal mechanisms, similar to the findings of Macias and colleagues (9, 15). A possible reason for the difference in the CVVH clearance of vancomycin between studies may pertain to differences in the applied hemofiltration rates. Also, existing literature varies in the distribution of pre-dilution and post-dilution replacement fluid administration, which is known to influence CVVH clearance of vancomycin(29). The present study elaborates on the unique attributes of critically ill patients exposed to concurrent vancomycin and CVVH. Distinct similarities exist with prior literature, particularly with respect to dialyzer selection and blood flow rate. Additionally, we document similarly poor trough achievement at 37% when compared to recently released literature on this topic (30-50%)(21, 33). The current analysis included a heterogeneous group of critically ill individuals exposed to a consistent weight-based dosing strategy with serum trough concentrations drawn under steady-

Page 9 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 state conditions. Also hemofiltration rates were generally higher than in previous studies and pre-dilution consistently comprised 50% of replacement fluid administration. Several studies note that increased doses of CRRT are not associated with improvement in patient outcomes(4, 6, 19, 32). The mechanism by which this treatment modality fails to provide benefit has yet to be fully understood. It is well known that early appropriate antibiotics are essential to the management of individuals with severe sepsis or septic shock and that delays in therapy are associated with dramatic increases in mortality(13). Appropriateness of antimicrobials encompasses not only spectrum of activity, but also adequacy of dosing with pharmacokinetic and pharmacodynamic optimization(10, 16). Given the present findings, it is possible that the absence of benefit associated with increasing hemofiltration rates during CRRT may be, in part, attributable to a reduced ability to achieve therapeutic goals for antimicrobials; however, further research is necessary to investigate this hypothesis. Several important limitations exist for this study. The retrospective nature limits the ability to perform formal pharmacokinetic modeling and establish causality. A potential selection bias exists due to the large portion of patients excluded for the absence of an available trough concentration. Although the strict study definition for trough concentrations led to reductions in the sample size, it facilitated assessment of concentrations under steady-state conditions which strengthens the analysis. Patient size, sex, and baseline serum-creatinine were found to be independently associated with prescribed hemofiltration rate. To our knowledge, this is the first time patient-specific factors which may influence dose prescribing practices have been characterized(28). Unfortunately, the only way to minimize confounding factors such as these would be to perform a randomized controlled trial which prospectively accounts for these attributes among hemofiltration rate dose groups. Vancomycin dose variability existed, however, the relative contribution of this factor to the outcome is likely minimal given the lack

Page 10 238 239 240 241 242 243 244 245 246 247 248 249 250 of a significant difference between groups at baseline. Finally, external validity of the results may be affected by utilization of different CRRT modalities and institutional CVVH practices, including, but not limited to, hemofiltration rate prescription patterns, blood flow rate and dialyzer selection. These study results suggest that critically ill patients exposed to current CVVH hemofiltration rates during vancomycin therapy may experience a significant reduction in the ability to achieve goal trough concentrations. Prospective studies are necessary in critically ill patients exposed to CVVH to determine the optimal vancomycin dosing strategy for attainment of goal trough concentrations. Acknowledgments Support for this project was provided by NIH/NCRR CTSA Grant Number UL1 RR024150. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. 251

Page 11 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 References 1. Ali, T., I. Khan, W. Simpson, G. Prescott, J. Townend, W. Smith, and A. Macleod. 2007. Incidence and outcomes in acute kidney injury: a comprehensive population-based study. J Am Soc Nephrol 18:1292-8. 2. Angus, D. C., W. T. Linde-Zwirble, J. Lidicker, G. Clermont, J. Carcillo, and M. R. Pinsky. 2001. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 29:1303-10. 3. Bagshaw, S. M., C. George, and R. Bellomo. 2008. A comparison of the RIFLE and AKIN criteria for acute kidney injury in critically ill patients. Nephrol Dial Transplant 23:1569-74. 4. Bellomo, R., A. Cass, L. Cole, S. Finfer, M. Gallagher, S. Lo, C. McArthur, S. McGuinness, J. Myburgh, R. Norton, C. Scheinkestel, and S. Su. 2009. Intensity of continuous renal-replacement therapy in critically ill patients. N Engl J Med 361:1627-38. 5. Boereboom, F. T., F. F. Ververs, P. J. Blankestijn, T. J. Savelkoul, and A. van Dijk. 1999. Vancomycin clearance during continuous venovenous haemofiltration in critically ill patients. Intensive Care Med 25:1100-4. 6. Bouman, C. S., H. M. Oudemans-Van Straaten, J. G. Tijssen, D. F. Zandstra, and J. Kesecioglu. 2002. Effects of early high-volume continuous venovenous hemofiltration on survival and recovery of renal function in intensive care patients with acute renal failure: a prospective, randomized trial. Crit Care Med 30:2205-11. 7. Boussekey, N., A. Chiche, K. Faure, P. Devos, B. Guery, T. d'escrivan, H. Georges, and O. Leroy. 2008. A pilot randomized study comparing high and low volume hemofiltration on vasopressor use in septic shock. Intensive Care Med 34:1646-53. 8. Bressolle, F., J. M. Kinowski, J. E. de la Coussaye, N. Wynn, J. J. Eledjam, and M. Galtier. 1994. Clinical pharmacokinetics during continuous haemofiltration. Clin Pharmacokinet 26:457-71. 9. Chaijamorn, W., A. Jitsurong, K. Wiwattanawongsa, U. Wanakamanee, and P. Dandecha. 2011. Vancomycin clearance during continuous venovenous haemofiltration in critically ill patients. Int J Antimicrob Agents 38:152-6. 10. Dellinger, R. P., M. M. Levy, J. M. Carlet, J. Bion, M. M. Parker, R. Jaeschke, K. Reinhart, D. C. Angus, C. Brun-Buisson, R. Beale, T. Calandra, J. F. Dhainaut, H. Gerlach, M. Harvey, J. J. Marini, J. Marshall, M. Ranieri, G. Ramsay, J. Sevransky, B. T. Thompson, S. Townsend, J. S. Vender, J. L. Zimmerman, and J. L. Vincent. 2008. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Crit Care Med 36:296-327. 11. Jeffres, M. N., W. Isakow, J. A. Doherty, P. S. McKinnon, D. J. Ritchie, S. T. Micek, and M. H. Kollef. 2006. Predictors of mortality for methicillin-resistant Staphylococcus aureus health-care-associated pneumonia: specific evaluation of vancomycin pharmacokinetic indices. Chest 130:947-55. 12. Joy, M. S., G. R. Matzke, R. F. Frye, and P. M. Palevsky. 1998. Determinants of vancomycin clearance by continuous venovenous hemofiltration and continuous venovenous hemodialysis. Am J Kidney Dis 31:1019-27. 13. Kumar, A., D. Roberts, K. E. Wood, B. Light, J. E. Parrillo, S. Sharma, R. Suppes, D. Feinstein, S. Zanotti, L. Taiberg, D. Gurka, A. Kumar, and M. Cheang. 2006. Duration of hypotension before initiation of effective antimicrobial therapy is a critical determinant of survival in human septic shock. Crit Care Med 34:1589-1596.

Page 12 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 14. Liu, C., A. Bayer, S. E. Cosgrove, R. S. Daum, S. K. Fridkin, R. J. Gorwitz, S. L. Kaplan, A. W. Karchmer, D. P. Levine, B. E. Murray, J. R. M, D. A. Talan, and H. F. Chambers. 2011. Clinical practice guidelines by the infectious diseases society of america for the treatment of methicillin-resistant Staphylococcus aureus infections in adults and children: executive summary. Clin Infect Dis 52:285-92. 15. Macias, W. L., B. A. Mueller, and S. K. Scarim. 1991. Vancomycin pharmacokinetics in acute renal failure: preservation of nonrenal clearance. Clin Pharmacol Ther 50:688-94. 16. Mueller, B. A., D. A. Pasko, and K. M. Sowinski. 2003. Higher renal replacement therapy dose delivery influences on drug therapy. Artif Organs 27:808-14. 17. O'Reilly, P., and A. Tolwani. 2005. Renal replacement therapy III: IHD, CRRT, SLED. Crit Care Clin 21:367-78. 18. Overberger, P., M. Pesacreta, and P. M. Palevsky. 2007. Management of renal replacement therapy in acute kidney injury: a survey of practitioner prescribing practices. Clin J Am Soc Nephrol 2:623-30. 19. Palevsky, P. M., J. H. Zhang, T. Z. O'Connor, G. M. Chertow, S. T. Crowley, D. Choudhury, K. Finkel, J. A. Kellum, E. Paganini, R. M. Schein, M. W. Smith, K. M. Swanson, B. T. Thompson, A. Vijayan, S. Watnick, R. A. Star, and P. Peduzzi. 2008. Intensity of renal support in critically ill patients with acute kidney injury. N Engl J Med 359:7-20. 20. RENAL Study Investigators. 2008. Renal replacement therapy for acute kidney injury in Australian and New Zealand intensive care units: a practice survey. Crit Care Resusc 10:225-30. 21. Roberts, D. M., J. A. Roberts, M. S. Roberts, X. Liu, P. Nair, L. Cole, J. Lipman, and R. Bellomo. 2012. Variability of antibiotic concentrations in critically ill patients receiving continuous renal replacement therapy: a multicentre pharmacokinetic study. Critical care medicine 40:1523-8. 22. Ronco, C., R. Bellomo, P. Homel, A. Brendolan, M. Dan, P. Piccinni, and G. La Greca. 2000. Effects of different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective randomised trial. Lancet 356:26-30. 23. Rybak, M. J., B. M. Lomaestro, J. C. Rotschafer, R. C. Moellering, W. A. Craig, M. Billeter, J. R. Dalovisio, and D. P. Levine. 2009. Vancomycin therapeutic guidelines: a summary of consensus recommendations from the infectious diseases Society of America, the American Society of Health-System Pharmacists, and the Society of Infectious Diseases Pharmacists. Clin Infect Dis 49:325-7. 24. Shen, H. N., C. L. Lu, and H. H. Yang. 2010. Epidemiologic trend of severe sepsis in Taiwan from 1997 through 2006. Chest 138:298-304. 25. Shorr, A. F., S. T. Micek, W. L. Jackson, Jr., and M. H. Kollef. 2007. Economic implications of an evidence-based sepsis protocol: can we improve outcomes and lower costs? Crit Care Med 35:1257-62. 26. Tolwani, A. J., R. C. Campbell, B. S. Stofan, K. R. Lai, R. A. Oster, and K. M. Wille. 2008. Standard versus high-dose CVVHDF for ICU-related acute renal failure. J Am Soc Nephrol 19:1233-8. 27. Trotman, R. L., J. C. Williamson, D. M. Shoemaker, and W. L. Salzer. 2005. Antibiotic dosing in critically ill adult patients receiving continuous renal replacement therapy. Clin Infect Dis 41:1159-66. 28. Uchino, S., R. Bellomo, H. Morimatsu, S. Morgera, M. Schetz, I. Tan, C. Bouman, E. Macedo, N. Gibney, A. Tolwani, H. Oudemans-van Straaten, C. Ronco, and J. A.

Page 13 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 Kellum. 2007. Continuous renal replacement therapy: a worldwide practice survey. The beginning and ending supportive therapy for the kidney (B.E.S.T. kidney) investigators. Intensive Care Med 33:1563-70. 29. Uchino, S., L. Cole, H. Morimatsu, D. Goldsmith, and R. Bellomo. 2002. Clearance of vancomycin during high-volume haemofiltration: impact of pre-dilution. Intensive Care Med 28:1664-7. 30. Uchino, S., N. Fealy, I. Baldwin, H. Morimatsu, and R. Bellomo. 2003. Continuous is not continuous: the incidence and impact of circuit "down-time" on uraemic control during continuous veno-venous haemofiltration. Intensive Care Med 29:575-8. 31. Uchino, S., J. A. Kellum, R. Bellomo, G. S. Doig, H. Morimatsu, S. Morgera, M. Schetz, I. Tan, C. Bouman, E. Macedo, N. Gibney, A. Tolwani, and C. Ronco. 2005. Acute renal failure in critically ill patients: a multinational, multicenter study. JAMA 294:813-8. 32. Van Wert, R., J. O. Friedrich, D. C. Scales, R. Wald, and N. K. Adhikari. 2010. High-dose renal replacement therapy for acute kidney injury: Systematic review and meta-analysis. Crit Care Med 38:1360-9. 33. Wilson, F. P., and J. S. Berns. 2012. Vancomycin levels are frequently subtherapeutic during continuous venovenous hemodialysis (CVVHD). Clinical nephrology 77:329-31. 367

Page 14 368 Table 1. Baseline patient characteristics and demographic data Characteristic HFR 30mL/kg/h c (N=10) HFR > 30mL/kg/h c (N=77) P Value Age (years) 68 (59-71) 62 (54-74) 0.55 Male (N; %) 5 (50) 51 (66) 0.32 APACHE II 26 (20-34) 25 (16-35) 0.99 SOFA 13 (6-16) 12 (7-15) 0.70 Body mass index (kg/m 2 ) 42 (31-55) 27 (24-32) < 0.001 Change in body weight (kg) a 1.2 (-3.9-5.1) 3.1 (0-9.4) 0.35 Net fluid balance (L) a 9.6 (-0.5-17.0) 10.3 (3.3-19.0) 0.59 ICU (N; %) Medical ICU Surgical ICU Other ICU 3 (30) 6 (60) 1 (10) 20 (26) 31 (40) 26 (34) 0.29 Admitting diagnosis (N; %) Non-operative Operative 5 (50) 5 (50) 41 (53) 36 (47) Sepsis (N; %) 9 (90) 69 (90) 1.00 Infection sources (N; %) b Respiratory Intra-abdominal Bacteremia Cardiovascular Musculoskeletal Other/Unknown 4 (40) 5 (50) 3 (30) 0 (0) 0 (0) 1 (10) 40 (52) 14 (18) 19 (25) 4 (5) 5 (6) 18 (23) 1.00 0.52 0.037 0.71 1.00 1.00 0.68 Vancomycin dose (mg/kg) 16.9 (14.8-20.0) 16.3 (15.3-19.7) 0.76 Target trough (N; %) 10-15 mg/l 15-20 mg/l 5 (50) 5 (50) 23 (30) 54 (70) 0.28 HFR: Hemofiltration rate; APACHE: Acute Physiology and Chronic Health Evaluation score; SOFA: Sequential Organ Failure Assessment score; ICU: intensive care unit a : From ICU admission to trough concentration b : Percentages may equal > 100 due to a patient having multiple infectious sources c : Values expressed as medians (interquartile ranges) unless noted 369