Infectious Diseases

Evaluation of the Chronic Kidney Disease Epidemiology
Collaboration Equation for Dosing Antimicrob...
KA Wargo and TM English

these equations estimate GFR and not CrCl, questions still                     proved observation...
Evaluation of Equation for Dosing Antimicrobials

Results                                                                 ...
KA Wargo and TM English

vance. Ideally, clinical data such as urine output, nutrition-            age adjustments to be s...
Evaluation of Equation for Dosing Antimicrobials

ist and merit further consideration, considering that 1 in 4            ...
KA Wargo and TM English

assessing outcomes, clinical significance can only be im-                 analysis and the Levey ...
Evaluation of Equation for Dosing Antimicrobials

tients.”17 The question should not be which equation we                 ...
KA Wargo and TM English

L’Evaluation d’Une Nouvelle Equation pour Estimer le Taux de              hospitalier de soins te...
Upcoming SlideShare
Loading in …5

Evaluation of the chronic kidney disease epidemiology 2010


Published on

Published in: Education
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Evaluation of the chronic kidney disease epidemiology 2010

  1. 1. Infectious Diseases Evaluation of the Chronic Kidney Disease Epidemiology Collaboration Equation for Dosing Antimicrobials Kurt A Wargo and Thomas M English lomerular filtration rate (GFR) is G the most accurate measure of one of the major functions of the kidneys BACKGROUND: Since the derivation of the Modification of Diet in Renal Disease (MDRD) equation for estimating glomerular filtration rate (GFR), investigators (clearance), though difficult and costly to determined that it cannot be used for drug dosing. In 2009, the Chronic Kidney directly measure.1 Clinically, estimation of Disease Epidemiology Collaboration (CKD-EPI) derived an equation that was more accurate than the MDRD estimation of GFR. Therefore, questions exist GFR using the Modification of Diet in Re- about which method should be preferred in making dosage adjustments for nal Disease (MDRD) equation allows renally eliminated antimicrobials. practitioners to stage chronic kidney dis- OBJECTIVE: To determine whether a difference exists when making antimicrobial ease.2 When making estimations of kidney dosage adjustments in patients with CKD based on estimation of GFR using the function at the bedside, for the purpose of CKD-EPI and Cockcroft-Gault equations. drug dosing, practitioners utilize the Cock- METHODS: A database of 409 patients with CKD admitted to a tertiary care facility croft-Gault equation.3 This equation esti- was used. GFR was calculated using both the CKD-EPI equation(s) and the mates creatinine clearance (CrCl) and is Cockcroft-Gault equation and compared using correlation and Bland-Altman recommended by the Food and Drug Ad- methodology. Dosage discordance rates of antimicrobials were determined. ministration (FDA) for use by pharmaceu- RESULTS: Average GFRs for all patients using the Cockcroft-Gault and CKD-EPI equations were 34.8 ± 12 mL/min and 39.9 ± 13 mL/min, respectively (5.09 [95% CI tical companies when specific renal dos- 4.60 to 5.59]; p < 0.001). The correlation coefficient between the 2 estimations was age adjustments are required.4 Since the high (r = 0.91). The Bland-Altman plot yielded limits of agreement of 15.3 and –5.1; derivation of the MDRD equation and the thus, the CKD-EPI estimation may range from 5.1 mL/min below to 15.3 mL/min finding that it was a more accurate predic- above the Cockcroft-Gault estimation for 95% of the cases. A discordance rate of tor of renal function than the Cockcroft- 15–25% existed among the recommended dosing adjustments of the selected Gault equation estimate, questions have antimicrobials when comparing the Cockcroft-Gault and CKD-EPI estimations. existed with regard to its use for the pur- CONCLUSIONS: Though this study did not determine which equation should be selected to dose adjust antimicrobials, it demonstrated statistically significant pose of making drug dosage adjustments.2 differences between the Cockcroft-Gault and CKD-EPI equations. The clinical To complicate matters, in 2009 a new set significance of these differences is uncertain in the absence of data assessing of equations for the measurement of GFR, clinical outcomes that result from the use of the discordant doses. Clinical derived from the Chronic Kidney Disease judgment should be employed when making renal dosage adjustments of Epidemiology Collaboration (CKD -EPI) antimicrobials. study, were found to provide a statistically KEY WORDS: antimicrobials, chronic kidney disease, CKD-EPI, Cockcroft-Gault, more accurate estimation of GFR than the dosing, MDRD. MDRD equation.5 Unfortunately, because Ann Pharmacother 2010;44:439-46. Published Online, 17 Feb 2010,, DOI 10.1345/aph.1M602 Author information provided at end of text. The Annals of Pharmacotherapy I 2010 March, Volume 44 I 439
  2. 2. KA Wargo and TM English these equations estimate GFR and not CrCl, questions still proved observational analysis conducted at an 881-bed ter- exist as to which method to use when estimating renal func- tiary care facility. A search engine was used to identify pa- tion for the purposes of drug dosing in patients with CKD. tients admitted with a SCr of 1.3–3 mg/dL. Patients were in- Differences exist between the CKD -EPI, MDRD, and cluded in the analysis if they were identified as having CKD Cockcroft-Gault estimations of kidney function.2,3,5 One dif- by physician documentation and were classified as CKD ference between the equations is that the 6-variable MDRD stages 3 (GFR 30–59 mL/min), 4 (15–29 mL/min), or 5 equation takes into account 3 biochemical markers, serum (<15 mL/min), using the MDRD equations. Excluded pa- creatinine (SCr), serum albumin, and blood urea nitrogen tients were those with acute renal dysfunction, defined as an (BUN), along with age, race, and sex. On the other hand, 8 elevation in SCr of 0.5 mg/dL from baseline, or from physi- CKD -EPI equations exist, which take into account race, sex, cian documentation, end-stage renal disease on dialysis, and SCr. The Cockcroft-Gault equation, on the other hand, is CKD stages 1 or 2, and those who were of a race other dependent only on weight and SCr (Table 1). than white or African American. One of the major responsibilities of pharmacists in- Estimation of GFR was performed using the CKD -EPI volves making drug dosing adjustments based on estima- equations and MDRD equations normalized to body sur- tions of renal clearance of medications. Therefore, it is in- face area (BSA) in order to determine the patient-specific creasingly important that an equation that accurately esti- GFR in milliliters/minute.8 The Cockcroft-Gault estima- mates this clearance is utilized when providing the most tion of renal function was used as the comparator equation, optimal drug dosing recommendations. While both the in which the lower of actual or ideal body weight (IBW) CKD -EPI and MDRD equations appear to more accurate- was used. For patients whose actual body weight exceeded ly estimate GFR than does the Cockcroft-Gault equation, their IBW by greater than 30%, an adjusted weight was they have not been validated for the purposes of making used in the calculation. Adjusted body weight was deter- drug dosage adjustments. Numerous research studies have mined by the equation [(actual body weight – IBW) ≥ 0.4] indicated that significant differences exist when comparing + IBW. Dosing discordance rates between the Cockcroft- Cockcroft-Gault and MDRD equations for estimating re- Gault and CKD -EPI equations were determined based on nal function for the purposes of making dosage adjust- the manufacturers’ renal dosing recommendations of 8 ments; however, this is the first study to compare the common antimicrobials (Table 2).7,9 These antimicrobials Cockcroft-Gault and CKD -EPI equations for this intent.6 were selected because the manufacturers’ recommenda- tions for dosage adjustment in renal dysfunction were Methods based on the Cockcroft-Gault equation. The methods for this study have been previously de- scribed.7 In short, this was an institutional review board–ap- STATISTICAL ANALYSIS Data were compiled in Microsoft Access (Microsoft Corp., Redmond, WA) and statisti- cal testing was completed using SPSS soft- Table 1. CKD-EPI Equations5 ware.10 Using the single proportion sample Serum size measurement, a total of 247 patients were Creatinine Race and Sex (mg/dL) Equation needed to detect a 20% discordance rate with a Black 95% confidence interval. Comparison of con- female ≤0.7 GFR = 166 × (SCr/0.7)0.329 × (0.993)Age tinuous variables was performed by using >0.7 GFR = 166 × (SCr/0.7)1.209 × (0.993)Age paired t-test and dichotomous variables were male ≤0.9 GFR = 163 × (SCr/0.9)0.411 × (0.993)Age compared using the χ2 test, as appropriate. >0.9 GFR = 163 × (SCr/0.9)1.209 × (0.993)Age Linear regression was incorporated to evaluate White or other correlations between continuous variables, as female ≤0.7 GFR = 144 × (SCr/0.7)0.329 × (0.993)Age appropriate. The Bland-Altman method was >0.7 GFR = 144 × (SCr/0.7)1.209 × (0.993)Age used to assess agreement between the CKD - male ≤0.9 GFR = 141 × (SCr/0.9)0.411 × (0.993)Age EPI and Cockcroft-Gault estimations of renal >0.9 GFR = 141 × (SCr/0.9)1.209 × (0.993)Age function.11,12 χ2 Analysis was used to detect a Single Equation GFR = 141 × min(SCr/k,1)α × max(SCr/k,1)–1.209 × 0.993Age × [1.018 if female] × [1.159 if black], where α = –0.329 for difference in dosing discordance data. Level of females and –0.411 for males; k = 0.7 for females and significance was set as p < 0.05. Data are pre- 0.9 for males; max = maximum of SCr/k or 1; and min = minimum of SCr/k or 1. sented as means (range) for continuous vari- ables and as a number for dichotomous vari- CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; GFR = ables, and 95% confidence intervals are report- glomerular filtration rate; SCr = serum creatinine. ed as appropriate. 440 I The Annals of Pharmacotherapy I 2010 March, Volume 44
  3. 3. Evaluation of Equation for Dosing Antimicrobials Results equation was 39.9 ± 13 mL/min (Table 4). The absolute mean difference between the 2 estimations was 5.10 ± 3.61 A total of 409 patients were eligible for evaluation in mL/min (95% CI 4.60 to 5.59; p < 0.001). Compared with this analysis, as previously described.7 The mean ± SD age our previous data, the absolute mean difference in GFR, of the cohort was 73.4 ± 12.5 years (Table 3). Patients using the MDRD and CKD -EPI equations, was 0.3 ± 0.22 were evenly distributed based on their sex, with the excep- mL/min (95% CI –0.14 to 0.76; p = 0.180). tion of SCr, which was higher in males than in females (p A correlation coefficient was determined for the rela- < 0.01). There was a preponderance of whites (81%) in the tionship of between calculated GFR using the CKD -EPI cohort studied. Mean weight was 80 ± 23 kg, BSA was and Cockcroft-Gault equations among the patients evaluat- 1.90 ± 0.26 m2, BUN was 35 ± 16 mg/dL, SCr was 1.75 ± ed. Excellent correlation existed among all patients (r = 0.5 mg/dL. Among the cohort of patients sampled, 46% 0.91); however, the line of unity demonstrated that CKD - weighed within 30% of their IBW, 36% exceeded their EPI estimations were consistently higher than Cockcroft- IBW by greater than 30%, and another 18% weighed less Gault estimations of GFR (Figure 1). When comparing the than their IBW. 2 estimates of GFR using the method described by Bland- When estimating renal function, the average CrCl, using Altman, the difference in values was plotted against the the Cockcroft-Gault equation, for all patients was 34.8 ± mean for the 2 methods in order to determine the variabili- 12 mL/min, whereas the average GFR using the CKD -EPI ty between them.10,11 The limits of agreement were 15.3 and –5.1; thus the CKD -EPI estimation may be 15.3 mL/min above or 5.1 mL/min below the Cockcroft-Gault estimation for 95% of the cases (Figure 2). For the upper Table 2. Manufacturer-Recommended Renal Dosing for limit of agreement, the confidence interval was 14.4 to Selected Antimicrobials7,9 16.2 mL/min and for the lower limit, the confidence inter- FDA-Recommended val was –6.0 to – 4.7 mL/min. Antimicrobial CrCl (mL/min) to Adjust Dosage Antimicrobial dosage discordance rates were calculated Cefazolin 10–30 to evaluate the difference between the estimations of renal <10 function (Figure 3). It was determined that an overall dis- Cefepime 30–60 11–29 cordant rate of 15–25% existed between the recommended <11 dosing adjustments of the selected antimicrobials when Daptomycin <30 comparing the Cockcroft-Gault and CKD -EPI estima- Levofloxacin 20–49 tions. This discordant rate was lower than in our previous 10–19 study comparing the Cockcroft-Gault and MDRD estima- Meropenem 26–50 10–25 tions, which found a difference of 20–36%.7 When com- <10 paring the CKD -EPI and MDRD estimations, a 7–12% Piperacillin/tazobactam 20–40 discordant rate was present. For the comparison of Cock- <20 croft-Gault and CKD -EPI estimations, the majority Trimethoprim/sulfamethoxazole 15–30 (88–96%) of discordance occurred when the manufacturer CrCl = creatinine clearance; FDA = Food and Drug Administration. recommended a dosage adjustment; however, the dosage adjustment was deemed unnecessary, accord- ing to estimation by the CKD -EPI equation. Table 3. Demographics7 Discussion Female Male Parameter (n = 208) (n = 201) One well-recognized component of clinical Age, y, mean (range) 75 (31–102) 72 (33–94) pharmacy involves renal dose adjustment of African American, n 45 33 pharmacotherapy, when deemed appropriate. SCr, mg/dL (mean ± SD) 1.66 ± 0.43 1.85 ± 0.55a The Pharmacy and Therapeutics committee at BUN, mg/dL (mean ± SD) 34.6 ± 16.6 35.4 ± 15.2 our institution has approved the right of clini- Actual weight, kg, mean (range) 77 (32–160) 83 (44–177) cal pharmacists to dose-adjust medications Ideal weight, kg, mean (range) 55 (43–84) 73 (55–94) based on FDA-approved manufacturer recom- Height, inches, mean (range) 64 (50–75) 70 (63–79) Actual weight >130% of ideal weight, n 104 41a mendations. This dosing modification may be Actual weight <ideal weight, n 26 46a done by pharmacists without prior approval from the physician. Therefore, having an accu- BUN = blood urea nitrogen; SCr = serum creatinine. rate estimation of renal function is of the ut- a p < 0.05. most importance and has much clinical rele- The Annals of Pharmacotherapy I 2010 March, Volume 44 I 441
  4. 4. KA Wargo and TM English vance. Ideally, clinical data such as urine output, nutrition- age adjustments to be statistically greater with the MDRD al status, trends in SCr, and severity of illness would play a equation compared with the Cockcroft-Gault equation us- more significant role in the decision by clinical pharma- ing actual body weight and the Cockcroft-Gault equation cists to adjust doses as needed for renal function than the using IBW (88%, 85%, 82%; p < 0.001). The results of estimates provided by equations. In reality, though, the use this large-scale study further validate the authors’ original of equations to estimate renal function play a bigger role in 1999 findings that the MDRD equation is more accurate this decision. Therefore, there is a strong need for estima- than Cockcroft-Gault estimations.2 Further, the authors tions of renal function to be as accurate as possible and in suggest that, when the MDRD equation is normalized to accordance with the method used by pharmaceutical com- BSA, it may be used for the purposes of drug dosing. Un- panies in the development of renal dosing guidelines. fortunately, because of current FDA mandates, it is unrea- In order to facilitate this process, and in accordance with sonable for the pharmaceutical industry to review all of the Levey and colleagues2 data that demonstrated that their dosage recommendations based on the more accurate MDRD estimations were more accurate than Cockcroft- MDRD, and now CKD -EPI equations. Further, with an Gault estimations, our institution reports the 4-variable absolute difference in concordance of 3% between the MDRD GFR in the routine chemistry panel of all patients. MDRD and Cockcroft-Gault estimation, it can be argued However, there is a lack of literature to date that has evalu- that clinical significance does not differ between the 2. ated the clinical utility of the MDRD or CKD -EPI equa- Therefore, continued use of the Cockcroft-Gault estima- tions for making dosage adjustments of renally eliminated tion of renal function is rational. medications. Further, considering that the FDA continues The findings of the present analysis demonstrate the ex- to require that the pharmaceutical industry make dosage istence of a statistically significant difference when com- adjustment recommendations based on the Cockcroft- paring the CKD -EPI and Cockcroft-Gault estimations of Gault equation, it makes it difficult to advocate the use of renal function. Even though a strong correlation existed MDRD or CKD -EPI. Finally, no equation gives an accu- when evaluating our entire cohort of patients, the Bland- rate estimate of renal function in patients with fluctuating Altman method for assessing agreement demonstrated a SCr concentrations; therefore, decisions to dose-adjust wide variation between the 2 estimations. Interestingly, medications should not be made based on the results of a this variation was smaller than in our previous study com- calculation of GFR at 1 moment in time. paring the Cockcroft-Gault and MDRD equations, 20.6 The existing literature comparing Cockcroft-Gault and and –9.8, though estimation of GFR was still higher with MDRD estimations of kidney function demonstrates a sig- both of these equations compared with the Cockcroft- nificant discordance between the 2 estimations.5 Since our Gault estimation.7 Additionally, based on the method cho- original comparison, published in 2005, others have veri- sen to estimate GFR, antimicrobials in this analysis would fied the discordance that exists between the 2 estima- still have been dosed differently 15–25% of the time, albeit tions.5,13,14 In 2009, Levey and colleagues attempted to re- lower than our previous results comparing the MDRD and solve the question of discordance between the MDRD and Cockcroft-Gault equations (20 –36%).7 These data imply Cockcroft-Gault estimations by comparing their estima- that the CKD -EPI estimation is closer to the Cockcroft- tions with a directly measured GFR (using iothalamate Gault estimation of renal function than is the MDRD. clearance) in over 5500 patients.5 In their study, they found However, the possibility of clinically important differences concordance rates with manufacturer-recommended dos- between the Cockcroft-Gault and CKD -EPI equations ex- Table 4. Mean Difference in Cockcroft-Gault, CKD-EPI, and MDRD Equationsa,7 Cockcroft-Gault CKD-EPI GFR MDRD GFR (mL/min), (mL/min), (mL/min), Characteristic Mean ± SD Mean ± SD Mean ± SD p Valueb Overall (n = 409) 34.8 ± 12.0 39.9 ± 12.5 40.2 ± 12.2 <0.001 Female (n = 208) 30.3 ± 10.9 34.5 ± 10.6 34.9 ± 10.3 <0.001 Male (n = 201) 39.5 ± 11.3 45.5 ± 11.9 45.7 ± 11.5 <0.001 White (n = 331) 35.4 ± 12.0 39.8 ± 12.4 40.0 ± 12.0 <0.001 African American (n = 78) 32.3 ± 11.7 40.2 ± 13.2 40.8 ± 13.1 <0.001 CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; GFR = glomerular filtration rate; MDRD = Modification of Diet in Renal Disease. a Based on demographics. b Statistical significance existed when comparing Cockcroft-Gault with both the CKD-EPI and the MDRD equations; no significant differences were observed when comparing CKD-EPI and MDRD. 442 I The Annals of Pharmacotherapy I 2010 March, Volume 44
  5. 5. Evaluation of Equation for Dosing Antimicrobials ist and merit further consideration, considering that 1 in 4 when the CKD -EPI estimation was used, leading to the patients would have received a different dose of medica- potential for adverse reactions such as seizures, arrhyth- tion based on equation selected. mias, renal failure, gastrointestinal symptoms, and neuro- As stated previously, the majority of discordance existed muscular hypersensitivity. Although the potential for such when the manufacturer recommended a dosage adjustment adverse reactions is quite low and may not bear clinical according to the Cockcroft-Gault estimation, yet that par- significance, the variation between the 2 estimations was ticular level of dosage adjustment was unnecessary accord- so great (15.3 to –5.1 mL/min), a clinically significant dif- ing to GFR estimation by the CKD -EPI equation. Accord- ference may be implied. However, without actually admin- ing to this rationale, in patients with discordant dosage rec- istering antimicrobials to the patients, directly measuring ommendations, 88–96% would have been overdosed GFR, comparing that measurement to our estimations, and Figure 1. Comparison of CKD-EPI and Cockcroft-Gault estimations of renal function for the study population using correlation (N = 409). The line of unity demonstrates that CKD-EPI estimations of GFR are consistently higher than Cockcroft-Gault estimations in our population. CG = Cockcroft-Gault; CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; GFR = glomerular filtration rate. Figure 2. Comparison of CKD-EPI and Cockcroft-Gault estimations of renal function for the study population using Bland-Altman plot (N = 409). The limits of agreement demonstrate that CKD-EPI estimations of GFR are 15.3 mL/min above to 5.1 mL/min below Cockcroft-Gault estimations in 95% of cases. CG = Cockcroft-Gault; CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; GFR = glomerular filtration rate. The Annals of Pharmacotherapy I 2010 March, Volume 44 I 443
  6. 6. KA Wargo and TM English assessing outcomes, clinical significance can only be im- analysis and the Levey and colleagues study.5 Of note, the plied from this analysis. cohort of patients in the CKD -EPI study exhibited a mean Interestingly, the discordance between MDRD and age of 47 ± 15 (internal validation set) and 50 ± 15 years CKD -EPI dosing recommendations was quite small, rang- (external validation set), whereas our analysis consisted of ing from 7% to 17%, indicating that little difference would a significantly older population, mean of 73.4 ± 12.5 years. exist when making dosing recommendations based on ei- However, results from a study by Cirillo and colleagues ther the CKD -EPI or MDRD equations. This, however, is suggests this difference may not be significant, as they not entirely surprising given the data presented in the Lev- found the MDRD equation to be a more accurate predictor ey and colleagues study,5 as well as the data from this of GFR than the Cockcroft-Gault equation in older pa- study compared with our previous study.7 tients.16 This analysis contains various limitations, based on a A final limitation of this analysis lies within our method series of assumptions. Measurement of actual GFR was of selecting patients. While we were able to recruit more than not conducted on patients. Instead, we relied on the data a sufficient number of patients to power this analysis, we did presented in the Levey and colleagues CKD -EPI study not include patients with SCr less than 1.3 mg/dL with sub- to establish that GFR can be accurately estimated, using stantially decreased renal function or patients with SCr their equation.5 Thus, the major limitation of this analy- greater than 3 mg/dL yet not on dialysis. Therefore, all of the sis is associated with the comparison of 2 estimated val- possible patients with stages 3–5 CKD were not captured. ues. Furthermore, drug concentration monitoring was Taking into consideration the data from the present not performed during this analysis due to a lack of re- study, along with previously reported information, we sources. agree with the comments by Stevens and colleagues when Because the CKD -EPI equation was chosen as the com- they stated, “It is time to move beyond the focus on differ- parator estimator of renal function, it becomes important to ences among equations and towards a focus on using the control for patient demographic differences between this most accurate clinical data to improve the care of our pa- Figure 3. Antimicrobial dosage discordance rate when comparing the MDRD and CKD-EPI estimations of GFR with the manufacturer-recommended dosage adjustment using the Cockcroft-Gault estimation.7 CKD-EPI = Chronic Kidney Disease Epidemiology Collaboration; GFR = glomerular filtra- tion rate; MDRD = Modification of Diet in Renal Disease; Pip/Tazo = piperacillin/tazobactam; Trim/Sulfa = trimethoprim/sulfamethoxazole. Black bars = MDRD vs CKD-EPI. White bars = Cockcroft-Gault vs CKD-EPI. Dotted bars = Cockcroft-Gault vs MDRD. 444 I The Annals of Pharmacotherapy I 2010 March, Volume 44
  7. 7. Evaluation of Equation for Dosing Antimicrobials tients.”17 The question should not be which equation we 13. Gill J, Malyuk R, Djurdjev O, Levin A. Use of GFR equations to adjust drug doses in an elderly multi-ethnic group—a cautionary tale. Nephrol use to dose-adjust antimicrobials but rather, “Is this the Dial Transplant 2007;22:2894-9. DOI 10.1093/ndt/gfm289 only tool we need to use?”18 In our opinion, assessment of 14. Golik MV, Lawrence KR. Comparison of dosing recommendations for clinical information of our patients should be the lone fac- antimicrobial drugs based on two methods for assessing kidney function: tor when deciding to dose-adjust medications. Cockcroft-Gault and Modification of Diet in Renal Disease. Pharma- cotherapy 2008;28:1125-32. DOI 10.1592/phco.28.9.1125 Results from previous studies have raised questions in 15. Stevens LA, Nolin TD, Richardson MM, et al. Comparison of drug dos- the minds of clinicians as to whether the MDRD equation, ing recommendations based on measured GFR and kidney function esti- and now the CKD -EPI equation, should be the preferred mating equations. Am J Kidney Dis 2009;54:33- 42. DOI 10.1053/j.ajkd.2009.03.008 method to estimate renal function, in order to make critical 16. Cirillo M, Anastasio P, De Santo NG. Relationship of gender, age, and decisions about medication dosing. Though the results of body mass index to errors in predicted kidney function. Nephrol Dial this study show that the CKD -EPI equation is closer than Transplant 2005;20:1791-8. DOI 10.1093/ndt/gfh962 the MDRD equation to the Cockcroft-Gault estimations, 17. Stevens LA, Nolin T, Levey AS. In reply to ‘Estimated GFR for drug dosing: a bedside formula,’ ‘Drug dose adjustments in patients with re- statistically and potentially clinically significant differ- nal impairment,’ ‘Use of the MDRD study equation for drug dosing,’ ences still exist. It is our opinion that differences between and ‘Estimated GFR vs creatinine clearance for drug dosing.’ Am J Kid- estimations of renal function will always exist and, ulti- ney Dis 2009;54:985-6. DOI 10.1053/j.ajkd.2009.08.017 mately, when faced with the decision to adjust dosages, 18. Wargo KA. Clinical judgment: to dose adjust antimicrobials or not. Phar- macotherapy 2008;28:281e-3e. clinical judgment should prevail. Kurt A Wargo PharmD BCPS, Associate Clinical Professor, Harri- son School of Pharmacy, Auburn University, Auburn, AL Evaluación de la Ecuación del Chronic Kidney Disease Thomas M English PhD, University of Alabama at Birmingham; Huntsville Regional Medical Campus, Huntsville, AL Epidemiology Collaboration para Ajustes en Dosis de Agentes Reprints: Dr. Wargo, 301 Governors Dr. SW, Suite 385C1, Antimicrobiales Huntsville, AL, fax 256/551-4567, KA Wargo y TM English Financial disclosure: None reported Ann Pharmacother 2010;44:439- 46. References EXTRACTO TRASFONDO: Estudios realizados han determinado que la ecuación 1. Slikensen JR, Kasiske BL. Laboratory assessment of kidney disease: derivada del estudio de Modificación de Dieta en la Enfermedad Renal clearance, urinalysis, and kidney biopsy. In: Brenner BM, Levine SA, (MDRD) para estimar la tasa de filtración glomerular (GFR) no puede eds. Brenner & Rector’s: the kidney. 7th ed. Philadelphia, PA: WB Saun- ser utilizada para realizar ajustes en dosis en pacientes renales. En el año ders, 2004:1107-19. 2009, el Chronic Kidney Disease Epidemiology Collaboration (CKD - 2. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more ac- EPI) derivó una ecuación más precisa que la ecuación MDRD para curate method to estimate glomerular filtration rate from serum creati- estimar la GFR. No está claro cuál método debe utilizarse para realizar nine: a new prediction equation. Modification of Diet in Renal Disease ajustes en dosis de agentes antimicrobiales que son eliminados renalmente. Study Group. Ann Intern Med 1999;130:461-70. OBJETIVO: Determinar si existe diferencia al realizar ajustes en dosis de 3. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum agentes antimicrobiales en pacientes con enfermedad crónica del riñón creatinine. Nephron 1976;16:31- 41. (CKD) cuando se estima la GFR utilizando las ecuaciones CKD -EPI y 4. Food and Drug Administration. Guidance for industry: pharmacokinetics Cockroft-Gault (CG) in patients with impaired renal function—study design, data analysis, MÉTODOS: Se realizó un análisis de observación de 409 pacientes con and impact on dosing and labeling. Rockville, MD: US Department of CKD admitidos a una facilidad de cuidado terciario. Se estimó la GFR Health and Human Services, May 1998. utilizando la ecuación de CKD -EPI y se comparó con el estimado de 5. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate GFR calculado con la ecuación de CG utilizando análisis de correlación glomerular filtration rate. Ann Intern Med 2009;150:604-12. y el método Bland-Altman. Se determinó la diferencia en dosis de los agentes antimicrobiales seleccionados al utilizar los valores de GFR 6. Greenberg E, Saad N, Abraham T, Balmir E. Drug dosage adjustment calculados. using renal estimation equations: a review of the literature. Hosp Pharm RESULTADOS: La GFR promedio de los pacientes se calculó en 34.8 ± 12 2009;44:577-83. mL/min al utilizar la ecuación CG y 39.9 ± 13 mL/min al utilizar la 7. Wargo KA, Eiland EH III, Hamm W, English TM, Phillippe HM. Com- ecuación CKD -EPI (5.09; 95% CI 4.60 y 5.59, p < 0.001). El coeficiente parison of the Modification of Diet in Renal Disease and Cockcroft- de correlación entre ambos estimados fue alto (r = 0.91). Los límites de Gault equations for antimicrobial dosage adjustments. Ann Pharma- concordancia en la gráfica Bland-Altman fueron 15.3 y –5.1. El estimado cother 2006;40:1248-53. DOI 10.1345/aph.1G635 de GFR calculado con la ecuación de CKD -EPI pudiera estar entre 5.1 8. DuBois D, DuBois E. A formula to estimate the approximate surface mL/min por debajo y 15.3 mL/min por encima del estimado con la area if height and weight be known. Arch Intern Med 1916;17:863-71. ecuación CG en el 95% de los casos. Se observó una diferencia de 15 a 9. Lacy CF, Armstrong LL, Goldman MP, Lance LL. Drug information 25% en los ajustes de dosis recomendados de los agentes antimicrobiales. handbook. 13th ed. Hudson, OH: Lexi-Comp, Inc., 2005. CONCLUSIONES: Este estudio demostró diferencias significativas en los 10. SPSS for Windows, Rel. 15.0.0., 2006. Chicago, IL: SPSS Inc. ajustes en dosis de agentes antimicrobiales al utilizar las ecuaciones 11. Bland JM, Altman DG. Statistical methods for assessing agreement be- CKD -EPI y CG. Se desconoce el significado clínico de estas diferencias tween two methods of clinical measurement. Lancet 1986;1:307-10. ante la ausencia de datos que evalúen los resultados clínicos asociados con la diferencia de las dosis calculadas. Se debe utilizar el juicio clínico 12. Bland JM, Altman DG. Applying the right statistics: analyses of mea- al hacer ajustes en dosis de agentes antimicrobiales. surement studies. Ultrasound Obstet Gynecol 2003;22:85-93. DOI 10.1002/uog.122 Traducido por Astrid J García-Ortiz The Annals of Pharmacotherapy I 2010 March, Volume 44 I 445
  8. 8. KA Wargo and TM English L’Evaluation d’Une Nouvelle Equation pour Estimer le Taux de hospitalier de soins tertiaires. Le TFG était calculé par les 2 équations à Filtration Glomérulaire en Présence d’Insuffisance Rénale l’étude et les résultats évalués par des analyses de corrélation et une analyse comparative de Bland-Altman. Le taux de discordance des KA Wargo et TM English recommandations des ajustements posologiques dérivés de ces différents Ann Pharmacother 2010;44:439- 46. estimés du TFG était finalement déterminé. RÉSULTATS: Les valeurs moyennes de TFG étaient de 34.8 ± 12 mL/min et de 39.9 ± 13 mL/min pour les formules CG et CKD -EPI, respective- RÉSUMÉ ment (différence moyenne absolue 5.1; intervalle de confiance de 95% INTRODUCTION: Plusieurs recherches démontrent que l’équation MDRD 4.6 – 5.59, p < 0.001). Une très bonne corrélation entre les 2 estimés a (modification de la diète en présence de maladie rénale-Modification of été notée (r = 0.91). Selon les limites d’entente déterminées par l’analyse de Diet in Renal Disease) pour évaluer le taux de filtration glomérulaire Bland-Altman, les estimés du TFG obtenus avec la formule CKD -EPI (TFG) ne peut être utilisée pour guider les ajustements posologiques de pouvaient être, dans 95% des cas, inférieurs de 5.1 mL/min et supérieurs différents médicaments. En 2009, le groupe d’épidémiologie sur de 15.3 mL/min par rapport aux valeurs obtenues par l’équation CG. Un l’insuffisance rénale chronique (CKD -EPI) a proposé une méthode plus taux de discordance pouvait varier entre 15 et 25% au niveau des précise que l’équation MDRD pour l’estimation du TFG. Ce nouvel différentes recommandations d’ajustements posologiques basées sur les outil n’a toutefois pas été évalué dans un contexte d’estimation de la estimations de la fonction rénale à l’étude. fonction rénale et de recommandation pharmacothérapeutique. CONCLUSIONS: Cette étude a démontré une différence statistiquement OBJECTIF: L’objectif de cette étude est de déterminer s’il existe une significative entre les estimés du TFG obtenus par la formule CG et différence entre les recommandations d’ajustements posologiques des l’équation CKD -EPI. La significative clinique d’une telle différence antibiotiques en présence d’insuffisance rénale, lorsque ces demeure toutefois à être précisée. Un jugement clinique est donc nécessaire recommandations sont basées sur l’estimation du TFG dérivé de la lors d’ajustements posologiques d’antimicrobiens pour un patient chez formule Cockcroft-Gault (CG) et de l’équation CKD -EPI. qui l’estimation de la fonction rénale se fait par différentes méthodes. MÉTHODOLOGIE: Il s’agit d’une étude rétrospective ayant évalué le Traduit par Sylvie Robert dossier de 409 patients insuffisants rénaux admis dans un centre Full text access to The Annals of Pharmacotherapy is available to subscribers. Personal, Student, and Resident Online Subscriptions To access full text articles through The Annals Web site (, simply enter your customer number which appears on the mailing label, in both the user name and password boxes. The customer number appears in the top row of the label. It starts with the letters TP and includes the first group of numbers. For example, the highlighted portion of this label is the customer number. TP 34712 0111 1108 TP 34712 John Q Clinician, PharmD. 123 Main St. Cincinnati, OH 45678 You would enter TP34712 (without a space between the letters and numbers) as the user name and password. Institutional Online Subscriptions Subscriptions have automatic full text access based on the customer’s IP address. If you have not submitted your IP address, please contact customer service and provide your IP address and cus- tomer number. 446 I The Annals of Pharmacotherapy I 2010 March, Volume 44